article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
@xmath2 decay is a magnetic dipole radiative transition with photon energy @xmath8 about 114 mev , and a relatively large branching fraction of about 2% .
this is a transition between 1@xmath9 states of charmonium and its rate can be easily calculated in potential models .
the transition does not change a spatial part of the wave function and its matrix element in the leading approximation equals one . a simple calculation without relativistic corrections
gives the result@xcite @xmath10 .
it was expected that relativistic corrections are of order 20 - 30% , similarly to the case of the electric dipole transitions in charmonium .
therefore , it was surprising when in 1986 the crystal ball collaboration measured this branching fraction in the inclusive photon spectrum and obtained @xmath11@xcite .
there are a lot of theoretical predictions for this decay rate@xcite@xmath12@xcite , based on qcd sum rules , lattice qcd calculations and so on , but as a rule they give values approximately twice as large as the crystal ball result .
this puzzle remained for more than twenty years . only in 2009 the cleo collaboration published@xcite the result of a new measurement of this branching fraction using analysis of 12 exclusive decay modes of @xmath3 .
the obtained value @xmath13 is closer to theory . combining the crystal ball and cleo results ,
pdg@xcite got @xmath14 with a scale factor of 1.6 . in this work
we report a result of the new independent measurement .
the photon spectrum in @xmath15 decay can be written as@xcite @xmath16 here @xmath17 is the matrix element of the transition , @xmath18 , @xmath19 and @xmath20 are c - quark charge ( in electron charge units ) and mass while @xmath21 is a breit - wigner function taking into account a nonzero @xmath3 width .
the matrix element tends to unity when @xmath8 tends to zero and decreases slowly with the photon energy increase .
cleo found that the photon line shape of this transition is asymmetric , and a breit - wigner function alone provides a poor fit to data .
its modification with @xmath22 improves the fit around the peak , but gives a diverging tail at higher photon energies . to suppress this tail , cleo used in their fit @xmath23 with @xmath24 mev , explaining it by the overlap of the ground state wave functions
however , such a form of the matrix element is valid for the wave functions of the harmonic oscillator only . in all other potentials ,
@xmath25 dependence will be proportional to some negative degree of @xmath8 when @xmath8 tends to infinity .
we tried to fit the cleo data using another line shape : at photon energy @xmath8 near the resonance the decay probability @xmath26 is proportional to @xmath27 , but at higher energies the factor @xmath22 is replaced with @xmath8 .
we found that the function @xmath28 , where @xmath29 , is also suitable . here
the correction factor @xmath30 is a smooth function near the resonance .
results of fits with the cleo function and our function are shown in table [ ta1 ] .
one can see that results on the @xmath3 mass , width , and decay rate are close , and the confidence level of the fit with our function is also good .
therefore , we use the latter function in the analysis of our data .
the experiment was performed at the kedr detector@xcite of the vepp-4 m collider@xcite .
the collider operates with a peak luminosity of about @xmath31 @xmath32 near the @xmath0 peak energy .
luminosity is measured with single bremsstrahlung online and with small - angle bhabha scattering offline .
two methods of beam energy determination are used : resonant depolarization with accuracy of @xmath33 kev and ir - light compton backscattering with accuracy @xmath34 kev@xcite .
this analysis is based on a data sample of @xmath35 @xmath36 collected at the @xmath0 peak .
three @xmath0 scans were performed . using a measured beam energy spread as well as known@xcite @xmath37 and @xmath38
, we calculate the @xmath0 production cross section at the peak and get @xmath39 .
event selection was performed in two steps . at the first step ,
multihadron decays of @xmath0 were selected with the following cuts : total energy in the calorimeters is greater than 0.8 gev ; at least four clusters with the energy greater than 30 mev in the calorimeters are detected ; at least one central track in the drift chamber is reconstructed ; there are no muon tubes activated in the third layer of the muon system .
these cuts suppress background from the cosmic rays , beam - gas interactions and bhabha events . at the second step
, photons in these events were identified .
a photon is a cluster in the liquid krypton calorimeter which is not associated with the reconstructed tracks in the drift chamber and has no time - of - flight ( tof ) scintillation counters activated in front of it .
according to the simulation , the photon detection efficiency with the above mentioned cuts is about 28% and is almost constant in the investigated range of the photon energies .
the inclusive photon spectrum and a fit to our data is shown in fig .
[ fig5 ] .
the spectrum was fit with a sum of the signal having a shape @xmath40 , convolved with the calorimeter response function ( logarithmic normal distribution@xcite with @xmath41 mev at 110 mev and a=-0.33 ) , and background .
the background shape was taken in the form @xmath42 , where the first term describes `` fake '' photons appearing due to nuclear interactions of hadrons in the calorimeter and usually having energies less than 60 mev , the second term is the second - order polynomial describing photons arising mainly from @xmath43 decays , and @xmath44 is the spectrum of charged particles ( a charged particle can be also misidentified as a photon ) .
the breit - wigner function of the form @xmath45 , where @xmath46 , was used in the fit .
we also tried to fit our data using other line shapes .
we used @xmath21 alone , @xmath47 , the cleo function , and our function . again , as first noted by cleo , we found that the @xmath21 function alone gives a shifted value of @xmath3 mass compared to other functions : @xmath48 mev/@xmath5 .
the function @xmath27 leads to a large tail at higher photon energies , giving for the branching fraction @xmath49 ( here the decay probability function was integrated up to @xmath50 ) .
the last two functions give close fit results .
since the cleo function has an exponential factor , the result for the branching fraction with this function can be considered as a lower limit .
the difference between the results obtained with the two last functions is used to estimate a systematic error appearing due to the unknown line shape .
systematic errors of our measurements are listed in table [ ta3 ] .
the uncertainty of the @xmath3 width leads to an error , which we evaluated varying the @xmath3 width in the fit by 2.2 mev ( the current pdg error ) . a systematic error related to the background subtraction
was estimated by using in the fit a polynomial of the third order instead of the second - order one , varying ranges of the fit , and applying or not the tof veto in photon selection .
the error in the number of @xmath0 produced was evaluated using the known uncertainty of the luminosity measurement .
since the cluster multiplicity is different in the simulation and experimental photon spectrum in @xmath0 decays , the error due to the photon detection efficiency was estimated by changing by 25% weights of events with small @xmath51 and large @xmath52 track multiplicities , and taking different mc generators for the @xmath3 decays .
the calibration of the photon energy scale was performed using @xmath53 decays and @xmath54 transitions .
a new direct measurement of @xmath2 decay was performed .
we measured the @xmath3 mass , width , and branching fraction of @xmath2 decay .
the values of the branching fraction and @xmath3 mass are sensitive to the line shape of the photon spectrum and it should be taken into account during analysis .
our results on the @xmath3 mass and width are : before our experiment these parameters were measured in @xmath0 and @xmath57 meson decays , as well as in @xmath58 and @xmath59 collisions .
measurements of crystal ball , mark3 , bes , and kedr were performed using the radiative @xmath0 decays , therefore , a mass shift should be taken into account .
crystal ball and kedr made such a correction , but mark3 and bes did not .
therefore we believe that mark3 and bes results on the @xmath3 mass should be corrected by approximately 4 mev towards higher values .
et al_. , _ rev.mod.phys . _ * 80 * , 1161 ( 2008 ) .
et al_. , _ phys .
d _ * 34 * , 711 ( 1986 ) .
m. shifman , _ z. physik c _ * 4 * , 345 ( 1980 ) .
a.yu khodjamirian , _ sov .
phys . _ * 39 * , 614 ( 1984 ) .
beilin , a.v .
radyushkin , _ sov .
phys . _ * 45 * , 342 ( 1987 ) .
n. brambilla , yu jia , a. vairo , _ phys .
d _ * 73 * , 054005 ( 2006 ) .
j.j . dudek _
et al_. , _ phys .
d _ * 73 * , 074507 ( 2006 ) .
n. brambilla _ et al_. , arxiv:1010.5827 .
et al_. , _ phys .
lett . _ * 102 * , 011801 ( 2009 ) .
k. nakamura _
et al_. , _ j. phys .
_ g * 37 * , 075021 ( 2010 ) .
et al_. , _ nucl .
instr . and meth .
_ a * 478 * , 420 ( 2002 ) . v. anashin _
et al_. , stockholm 1998 , epac 98 * , 400 ( 1998 ) , prepared for 6th european particle accelerator conference(epac 98 ) , stockholm , sweden , 22 - 26 jun 1998 .
et al_. , _ nucl .
instr . and meth .
_ a * 598 * , 23 ( 2009 ) .
v.m . aulchenko _
et al_. , _ nucl . instr . and meth .
_ a * 379 * , 475 ( 1996 ) . | we present a study of the inclusive photon spectrum from 6.3 million @xmath0 decays collected with the kedr detector at the vepp-4 m @xmath1 collider .
we measure the branching fraction of the radiative decay @xmath2 , @xmath3 width and mass . taking into account an asymmetric photon line shape
we obtain : @xmath4 mev/@xmath5 , @xmath6 mev , @xmath7 . |
based on infinite product expansions and inequalities on series and the riemann s zeta function , daurizio ( @xcite ) proved the following inequality : @xmath0 where @xmath1 . using an elementary approach ,
sndor ( @xcite ) offered an alternative proof of by employing trigonometric inequalities and an auxiliary function . in the same paper ,
sndor also provided the converse to : @xmath2 where @xmath1 .
in addition , sndor found the following analogous inequality holds true for the case of sine functions : [ thm : daurizio - sandor ineq original ] the two double inequalities @xmath3 and @xmath4 hold for any @xmath1 . throughout this paper , we denote @xmath5 and @xmath6 by @xmath7 and @xmath8 , respectively : @xmath9 our aim is to generalize the daurizio - sndor inequalities for the case of @xmath10 and @xmath11 as follows : [ thm : daurizio - sandor ineq generalized ] let @xmath12 .
then the two double inequalities @xmath13 and @xmath14 hold for @xmath15 . in particular , the double inequality remains true when @xmath16 while the double inequality is reversed when @xmath16 .
the remainder of this paper is organized as follows .
section [ sec : proof of the main results ] is devoted to the proof of theorem [ thm : daurizio - sandor ineq generalized ] and an alternative proof of theorem [ thm : daurizio - sandor ineq original ] . in section [ sec : daurizio - sandor inequalities for hyperbolic functions ] , we establish analogue of theorem [ thm : daurizio - sandor ineq generalized ] for hyperbolic functions . as an application of theorem [ thm :
daurizio - sandor ineq generalized ] , we apply in section [ sec : application ] inequality to the chebyshev polynomials of the second kind and establish a trigonometric inequality .
at first we will prove the following lemma .
the lemma provides expressions of the higher - order derivative @xmath17 involving @xmath18 @xmath19 , which are helpful in proving theorem [ thm : daurizio - sandor ineq generalized ] .
we note that the sign of @xmath17 plays a crucial role in proving theorem [ thm : daurizio - sandor ineq generalized ] .
[ lem : lots of f ] let @xmath12 and @xmath20 .
then when @xmath21 and @xmath22 , we have @xmath23 : : @xmath24 @xmath25 : : @xmath26 + in particular , @xmath27 : : when @xmath28 , @xmath29 @xmath30 : : when @xmath31 , @xmath32 @xmath33 : : for @xmath34 and @xmath35 , @xmath36 @xmath23 , @xmath25 and @xmath33 follows directly from calculations using elementary calculus .
in particular , trigonometric addition formulas are used in proving @xmath23 and @xmath25 . to prove ,
we claim @xmath37 indeed , we rewrite @xmath38 on the other hand , making use of euler s formula @xmath39 leads to an alternative expression of the left - hand side of : @xmath40 where @xmath41 is the real part of @xmath42 and @xmath43 .
now it suffices to show @xmath44 using in @xmath25 , this can be achieved by straightforward calculations .
thus @xmath27 is true .
the proof of @xmath30 is similar , and we omit the details .
we complete the proof of lemma [ lem : lots of f ] .
we provide here an alternative proof of the two double inequalities in theorem [ thm : daurizio - sandor ineq original ] . to this end , we show that for @xmath1 , @xmath45 is strictly increasing while @xmath46 is strictly decreasing . these lead to the desired inequalities since it is easy to see that @xmath47 to see @xmath48 is strictly increasing , we employ in lemma [ lem : lots of f ] to obtain @xmath49 as @xmath50 , it follows that @xmath51 .
we are led to @xmath52 or @xmath53 since @xmath54 .
this that shows @xmath48 is strictly increasing . by using in lemma [ lem : lots of f ]
, we have @xmath55 from which we infer that @xmath56 since @xmath57 by @xmath33 of lemma [ lem : lots of f ] . then @xmath58 together with the fact @xmath59 from @xmath33 of lemma [ lem : lots of f ] yields @xmath60 or @xmath61 .
thus we have shown that @xmath62 is strictly decreasing .
this completes the proof of the theorem .
we are now in the position to give the proof of theorem [ thm : daurizio - sandor ineq generalized ] .
the proof of the case when @xmath16 has been given in theorem [ thm : daurizio - sandor ineq generalized ] . for @xmath63 , we prove the desired inequalities by showing that @xmath64 for @xmath34 .
due to @xmath23 of lemma [ lem : lots of f ] , we see that @xmath65 for @xmath63 . instead of employing @xmath25 of lemma [ lem : lots of f ] , we use and in lemma [ lem : lots of f ] to conclude that @xmath66 .
thus we have for @xmath34 , @xmath67 because of the first vanishing limit in @xmath33 of lemma [ lem : lots of f ] , it follows that @xmath68 which , together with the fact that the second limit in @xmath33 of lemma [ lem : lots of f ] vanishes , implies that @xmath69 or @xmath64 for @xmath34 .
it remains to find the following limits : @xmath70 we immediately have @xmath71 and @xmath72 the proof is completed .
in this section , we show an analogue of theorem [ thm : daurizio - sandor ineq generalized ] for the case of hyperbolic functions holds true .
let @xmath73 following the same arguments for proving lemma [ lem : lots of f ] , it can be shown that lemma [ lem : lots of f ] with @xmath74 , @xmath75 and @xmath18 @xmath19 replaced by @xmath76 , @xmath77 and @xmath78 @xmath19 respectively , remains true .
it follows that we can prove @xmath79 for @xmath34 as in the proof of theorem [ thm : daurizio - sandor ineq generalized ] .
it remains to calculate the following limits : @xmath80 thus , we have the following analogue of theorem [ thm : daurizio - sandor ineq generalized ] for @xmath76 and @xmath77 .
[ thm : daurizio - sandor ineq generalized hyperbolic fcn ] let @xmath12 .
then the two double inequalities @xmath81 and @xmath82 hold for @xmath15 . in particular , the double inequality is reversed when @xmath16 while the double inequality remains true when @xmath16 .
the first few chebyshev polynomials of the second kind @xmath83 @xmath84 are ( @xcite ) @xmath85 in this section , we apply theorem [ thm : daurizio - sandor ineq generalized ] to @xmath83 with @xmath86 . by means of the formula @xmath87 , we obtain the following corollary .
_ acknowledgements_. the authors wish to express sincere gratitude to tom mollee for his careful reading of the manuscript and valuable suggestions to improve the readability of the paper .
thanks are also due to chiun - chuan chen and mach nguyet minh for the fruitful discussions .
the authors are grateful to the anonymous referee for many helpful comments and valuable suggestions on this paper .
m. abramowitz and i. a. stegun ( eds ) handbook of mathematical functions with formulas , graphs , and mathematical tables national bureau of standards , applied mathematics series * 55 * , 9th printing washington 1970 | in this work , we generalize the daurizio - sndor inequalities ( @xcite ) using an elementary approach . in particular , our approach provides an alternative proof of the daurizio - sndor inequalities .
moreover , as an immediate consequence of the generalized daurizio - sndor inequalities , we establish the daurizio - sndor - type inequalities for hyperbolic functions . |
an extreme geomagnetic storm on 20 november 2003 was caused by the interaction of the earth s magnetosphere with an interplanetary magnetic cloud ( mc ) , whose magnetic helicity , @xmath0 , was positive . the very strong magnetic field in the mc of up to @xmath1 nt had a long - lasting negative @xmath2 component ( up to @xmath3 nt ) .
these circumstances were crucial in identifying the solar source for the mc . and definitely associated the mc with the filament eruption in the active region ( ar ) 10501 ( we will use henceforth the last three digits for brevity ) on 18 november .
these authors considered the direction of the axial magnetic field in the pre - eruptive filament to correspond to the expected projection @xmath4 .
with such a direction of the axial field , the current helicity of the filament , @xmath5 , is positive . from the condition @xmath6 , which is valid for a linear force - free magnetic field
, it follows that the magnetic helicity is also positive .
later found a correspondence between the flare reconnected magnetic flux , measured as the flare ribbon flux , and the poloidal magnetic flux of the mc under the assumption that the mc was a part of a magnetic flux rope with a length of 0.52 au .
the studies by , , and gave the impression of an acceptable correspondence between the conditions of the eruption in ar 501 and the parameters of the mc observed in the earth orbit : @xmath7 .
the study of changed this situation . from the observed morphological features they found that the large - scale magnetic field in ar 501 had a negative helicity sign .
this finding seemingly contradicted what was expected from the magnetic helicity conservation requiring the same sign of the magnetic helicity in the ar and mc .
this circumstance has raised a question , why the ar , which had a global negative magnetic helicity , could expel a positive - helicity mc to the interplanetary medium .
one possible answer was proposed by , who found a localized positive helicity injection in the southern part of ar 501 and concluded that the right handedness of the observed mc was due to the ejection from this portion of the ar . on the other hand , in their study of 12 interplanetary mcs and related solar active regions
have found : _ i _ ) a significant difference between the total twist of the magnetic field inside active regions , @xmath8 , and that in the mc , @xmath9 ; _ ii _ ) the absence of any significant sign relationship between them .
the authors used the linear force - free approximation , @xmath10 is a constant .
the dipole scale , @xmath11 , was measured as the distance between the centroids of the positive and negative fluxes in the magnetogram of an ar .
the magnetic field in an mc was fit with the lundquist magnetic model with @xmath12 au length .
findings ( _ i _ ) and ( _ ii _ ) have compelled to conclude that `` _ _ magnetic clouds associated with active region eruptions are formed by magnetic reconnection between these regions and their larger - scale surroundings , rather than simple eruption of preexisting structures in the corona or chromosphere _ _ '' .
described a similar phenomenon analytically by using the idealized example of the axially - symmetric reconfiguration of two twisted magnetic fluxes from their unconnected initial state to the connected relaxed state .
they have shown that magnetic reconnection can reverse the twist direction of a flux rope emerging into preexisting fields under the conservation of the total relative magnetic helicity .
the complex reconnection of a flux rope with the adjacent field in complex magnetic topology has been also described by , _
e.g. _ , , , and .
@xcite ( @xcite ) have found observational evidence of magnetic reconnection between the internal field belonging to the eruptive filament and the preexisting coronal field .
this is a phenomenon of plasma dispersal from an eruptive filament over the solar surface that is visible as the disintegration of the filament .
the whole mass of an eruptive filament or a considerable fraction of its mass does not leave the sun as a part of a cme .
the motion of the cool plasma of the eruptive filament continues along new magnetic field lines passing inside the eruptive filament and ending far on the solar surface .
clouds of such plasma can screen the emission of compact sources in active regions as well as the emission from quiet solar areas .
absorption phenomena can be observed in microwaves and also in the he ii 304 line .
events of such a kind are associated with active region eruptions .
they have been rarely detected for observational reasons .
the analysis of the solar geoeffective event of 18 november 2003 by ( hereafter paper i ) and ( hereafter paper ii ) has revealed that the major eruption in this event , _
i.e. _ , the eruption of the u - shaped filament , which we call f1 , from ar 501 , was also not a simple one . the eruptive filament bifurcated and transformed into a large y - shaped cloud , which moved from the region of bifurcation ( rb ) to the south
west across the solar disk toward the limb .
figure [ f - shape_transform ] illustrates what has happened to the eruptive filament , as shown by the h@xmath10 images produced by the kanzelhhe solar observatory ( kso ) , the _ extreme - ultraviolet imaging telescope _
( eit ; ) , on board the _ solar and heliospheric observatory _ ( soho ) , and the _ spectroheliographic x - ray imaging telescope _ ( spirit ; and ) , on board the _ complex orbital near - earth observations of activity of the sun _ ( coronas - f ) satellite .
line - center image ( kso ) .
the frames denote the fields of view of the four images shown in the lower row ( b e ) .
the axes are in arc seconds from the solar disk center .
the turquoise oval in all the images denotes the region rb where the filament bifurcated .
the yellow curves roughly outline the frontal edge of the filament before and during the eruption .
( c ) soho / eit image in the 195 channel .
( d ) kso h@xmath10 image observed in the far blue wing .
( e ) y - like cloud in the coronas - f / spirit 304 image . ]
the masses of the y - like cloud and the pre - eruption filament were similar ( paper i ) ; on the other hand , remnants of the filament were not evident in the southwestern cme observed by the _ large angle and spectrometric coronagraph _ ( lasco ) that was previously regarded as the source of the 20 november geomagnetic storm .
the observations analyzed in paper i and paper ii suggest a possible additional eruption in the interval from 08:07 to 08:14 ut above the bifurcation region close to the solar disk center that could be the source of the interplanetary mc on 20 november .
these facts disfavor the simple scenario , in which the 20 november mc is considered as a flux rope formed directly from a structure initially associated with the pre - eruptive filament f1 in region 501 .
as mentioned , the right - handed mc produced in this event and responsible for the superstorm had a very strong magnetic field near earth of up to @xmath1 nt and @xmath13 nt . according to , its inclination to the ecliptic plane was @xmath14 , and the magnetic flux in this plane was @xmath15 mx ; however , its significant part could be lost by reconnection in the interplanetary space . in ( hereafter paper iv )
, we additionally find that the mc was compact , with a size of about 0.2 au , and had some atypical properties , such as a wide range of proton temperatures , from @xmath16 k to @xmath17 k ; its magnetic structure was closed , disconnected from the sun , and probably had a spheromak configuration . the present paper ( hereafter paper iii ) endeavors to understand how the catastrophe of the eruptive filament could occur and create the right - handed spheromak - like mc .
section [ s - outline ] outlines the eruptive event and results of its analysis .
section [ s - helicity ] analyzes the helicity in ar 501 . in section [ s - large_scale_config ]
we address the causes of the bifurcation of the eruptive filament . in section [ s - mc_formation ]
we try to understand how the mc could be formed .
section [ s - summary ] briefly summarizes the outcome of this study .
paper i and paper ii have revealed a few eruptive episodes in the complex event of 18 november 2003 .
here we outline and illustrate them using the images from eit in figures [ f - history]a[f - history]f and the time profiles in figure [ f - history]g produced in hard x - rays ( hxr ) by the _
reuven ramaty high - energy solar spectroscopic imager _
( rhessi : ) and , for the rhessi nighttime , in microwaves by the us air force _ radio solar telescope network _ radiometers in san vito . the pre - eruptive u - shaped filament f1 rooted in ar 501 was pointing in the plane of the sky toward a small ring - like structure rb ( figure [ f - history]a ) .
active region 503 was located north of rb .
episode e1 presumably gave rise to a first southeastern cme ( cme1 ) responsible for am elongated dimming d1 , which started to develop during the interval shown in figure [ f - history]b .
a jet - like ejection , ej , in figure [ f - history]c ( e2 ) triggered the motion of filament f1 , but did not produce any cme directly .
the filament accelerated in a weak episode e3 ( figure [ f - history]g ) .
the eruptive filament , whose trajectory crossed a topological discontinuity located at a height of about 100 mm , collided with a coronal structure above rb ( e4a ) and bifurcated ( e4b ) , apparently rolling around it . in response to the interaction , brightenings overlapping rb appeared in figure [ f - history]d .
they rotated clockwise and , after e4b , vanished in figure [ f - history]e .
episode e4c was presumably related to the onset of the second southwestern cme ( cme2 ) . in figure [ f - history]e , dimmings developed at the previous position of rb and west of it ( d2 ) ; a central brightening ( not mentioned in papers i and ii ) appeared in ar 503 ( e4d ) , indicating its involvement in an eruption .
after the chain of eruptions e1e4 , regions d1 , d2 , d3 , and the former rb region dimmed ( figure [ f - history]f ) ; the bifurcated filament f1 apparently transformed into a y - like ` cloud ' moving across the solar disk ( figures [ f - shape_transform]c[f - shape_transform]e ) .
one more ejection probably occurred near the solar disk center during the e4 burst ( figures [ f - history]d , [ f - history]e ) .
the ejection was detected in paper ii as a faintly visible round feature , which expanded approximately from the position of the bifurcation region in the images produced by the _ solar x - ray imager _ ( sxi ) on board the _ geostationary operational environmental satellite _
( goes-12 ) .
the observed radial expansion was coordinated with the trajectory of a drifting type iv radio burst in the dynamic radio spectrum using the acceleration profile in figure [ f - history]h .
the final speed of this radial expansion was @xmath18 km s@xmath19 .
the mass of this ejection should be @xmath20 g ( see paper i ) .
the properties of the presumable ejection are very different from those of cme2 ejected at the same time , but appear to match the expectations for a source of the 20 november superstorm .
this low - mass , weakly expanding ejection presumably moved along the sun
earth line , and therefore its meager thomson - scattered light was insufficient to be detected by the lasco coronagraphs .
on the other hand , neither cme1 nor cme2 seem to be a promising candidate to be the source of the superstorm , being able to produce , at most , a glancing blow on the earth s magnetosphere ( paper ii ) . according to three reconstructions of the mc responsible for the geomagnetic superstorm @xcite , its dimensions in the ecliptic plane were @xmath21 au .
this size corresponds to an expansion angle of @xmath22 , which is similar to that of the presumable cme ( paper ii ) . to meet the earth , the mc expanding in such a narrow angle
should be ejected close to the solar disk center .
such a weak expansion is favored , if the mc is disconnected from the sun .
these speculations will be supported in paper iv .
the observational suggestions and the listed conjectures indicate that the ejection responsible for the superstorm probably originated during the e4 burst .
the formation of the ejection started at e4a at a height of about 100 mm ( @xmath23 ) and was completed by the end of e4d .
the latter is supported by the appearance of the central brightening in region ar 503 ( figure [ f - history]e ) .
a possible height , at which the formation was completed , is of the order of @xmath24 , with @xmath25 being the duration of burst e4 , and @xmath26 km s@xmath19 being an average sun earth transit speed of the icme ( estimated from the decrease of the dst index on 20 november ) .
this height can be overestimated by a factor of 23 due to the uncertainties in the velocity as the cme can accelerate as it leaves the sun and decelerate during its transit away from the sun .
a probable height , at which the formation was completed , is therefore between @xmath27 and @xmath28 .
it is difficult to expect to have direct observations of the processes , which occurred at the heights previously estimated close to the solar disk center .
therefore , we are forced to involve indirect observational indications and calculations .
we use the observational indication provided by the evolution of the dimming regions d1d3 in figure [ f - dimming ] .
their configuration is visible in the eit 195 difference image shown in figure [ f - dimming]a .
the time profiles of the average brightness in selected regions having the lowest intensity and longest lifetime are presented in figure [ f - dimming]b . unlike , who defined a dimmed region as the brightness decrease below the quiet - sun brightness level ( the horizontal line in figure [ f - dimming]b ) , we select the dimmings simply by a decrease of the brightness with respect to the pre - event level .
our different selection criterion indicates that we are considering a different kind of phenomenon . and @xmath29 .
the horizontal line presents the average quiet - sun intensity level in the eit 195 images , two of which were used for the difference in panel ( a ) . ]
most likely , the major cause of a dimming ( in our definition ) is the density decrease in the coronal structures due to their expansion .
the brightness in the extreme - ultraviolet ( euv ) and soft x - ray images is proportional to the column emission measure .
the brightness , @xmath30 , of an expanding closed coronal structure of a linear size , @xmath31 , and area , @xmath32 , filled with a fixed total number of emitting particles , @xmath33 , should be @xmath34 .
thus , the expansion alone should result in a considerable brightness decrease and a strong pressure gradient , which causes a secondary subsonic plasma outflow in the footprint regions of a cme @xcite .
on the one hand , the outflow is responsible for the redistribution of the coronal plasma from footprints into the expanding volume . on the other hand , due to the subsequent plasma supply from the chromosphere - to - corona transition region
, the outflow probably becomes the major factor to recover the brightness in the dimmed regions .
this simple consideration also explains why the development of dimming is often observed to start before the eruption , when coronal structures gradually expand during the initiation phase . the time profiles in figure [ f - dimming]b show that different parts of the star - like dimming d3 recovered considerably faster than the long - lived dimmings d1 and d2 .
the analysis of the magnetic connectivity shows that the plage region , where dimming d2 occurred , was connected to ar 503 associated with dimming d3 .
in addition , a fan of long diverging field lines rooted in the plage region connected it to remote regions far to the south west .
possibly , d2 shared a footprint of cme2 and had therefore the time profile typical of long - lived core dimmings . the pronouncedly shorter lifetimes of regions @xmath35 , @xmath29 , and 4 of the atypical dimming d3 hints at its involvement into an eruption , in which a structure disconnected from the sun developed .
the disconnection produced a stretched magnetic loop .
its subsequent evolution is determined by the relation between the magnetic tension , which tends to shrink the loop , and the opposite influence of the plasma pressure and the plasma outflow responsible for the dimming in its bases .
the shrinking duration for the loop with a length of @xmath36 mm is presumably comparable to the time required to stop the plasma outflow , @xmath37 s ( @xmath38 km s@xmath19 is a mean sound speed of the plasma outflow originating from the low corona up to the chromosphere - to - corona transition region ) .
this estimate is of the order of the observed formation time of dimming d3 .
figure [ f - suggestion ] summarizes the listed observational indications of a possible interaction between the eruptive filament ( red ) and a large static coronal structure ( blue ) .
one of its ends is rooted in ar 503 , and the opposite - polarity basis corresponds to a plage region and region of bifurcation , rb .
the onset of the interaction and bifurcation of the filament in shown in figure [ f - suggestion]a .
in a subsequent expansion of the eruptive filament in figure [ f - suggestion]b , its top brakes at the blue structure , while its lateral portions stretch out in a y - like form . in response to the interaction ,
dimmings develop in the plage region ( d2 ) and rb as well as dimming d3 at the opposite end of the blue structure rooted in ar 503 .
most likely , the eruptive filament was far from ar 503 ( _ cf .
_ figures [ f - history ] and [ f - dimming ] ) , which nevertheless appears to be implicated .
we shall use this presumable scheme as a hint in our considerations .
a major step in our study is the analysis of the coronal magnetic field _ via _ the extrapolation of photospheric magnetograms . for ar 501
we have a vector magnetogram that makes possible a non - linear force - free ( nlff ) field extrapolation within its field of view , @xmath39
. however , the region of interest , including rb , d2 , and d3 in figure [ f - dimming]a , is far west from the vector magnetogram .
the only possible way to analyze a larger region is a potential extrapolation of the full - disk soho / mdi magnetograms @xcite .
the potential approximation is usually considered to be insufficiently accurate to describe realistic configurations under the presence of significant electric currents . nevertheless , in paper i we have used it successfully to visualize the filament , its height , and the topological discontinuity , whose presence accounted for the apparent disintegration of the eruptive filament .
let us try to check , how realistic the results of the potential extrapolation are by comparing them with real coronal loops in euv images .
figure [ f - f_lines_center ] compares the field lines computed in the potential approximation , using the method and package of and for the full - disk soho / mdi magnetogram observed on 18 november at 09:35 ut . the spatial resolution is determined by 90 spherical harmonics .
the starting points for all of these field lines were chosen manually , and their density does not correspond to the real magnetic field strength distribution .
the correspondence between the computed field lines and the visible coronal loops is what is relevant to us .
figure [ f - f_lines_center]a presents the eit 195 image observed before the event in gray scale overlaid with a few sets of computed magnetic field lines .
the same sets of the field lines are overlaid on the post - event eit 284 image in figure [ f - f_lines_center]b .
the red lines embrace the pre - eruptive filament f1 .
the light - blue lines correspond to open magnetic field lines rooted in the coronal hole east of ar 501 .
the green lines cover a southern filament channel ; the corresponding coronal loops appeared after the event in figure [ f - f_lines_center]b .
all other structures are traced by the blue lines .
although a one - to - one correspondence is not observed , as expected , the figure shows an acceptable similarity of the computed blue field lines to the structures in the euv images .
the pre - event eit 195 image in figure [ f - f_lines_center]a presents rather short loops reaching small heights .
the computed field lines correspond to the loops diverging from ar 503 and from the plage region slightly south west of rb .
the overall correspondence in shapes and directions between the computed and real structures is observed to the south east and north east of ar 501 .
the red field lines above the filament correspond to magnetic structures , which are expected to be seen but are not visible .
these lines fairly correspond to the post - eruption arcade to the south of ar 501 in figure [ f - f_lines_center]b .
the light - blue open field lines are not expected to be visible .
the post - event eit 284 image in figure [ f - f_lines_center]b shows somewhat higher loops . for example , connections between ar 501 and ar 503 become detectable here .
in addition to the mentioned features , note that region rb is not seen here .
dimming d2 resembling a transient coronal hole has decreased in size with respect to its appearance in figure [ f - dimming ] .
we can not observe in these images the structures , which reach still larger heights , because of their lower brightness ( due to a lower plasma density ) against the brighter lower loops .
they can only be detected against the sky on the limb .
however , the on - limb magnetogram is very uncertain .
we therefore have computed the field lines extrapolated from the magnetogram observed on 18 november at 09:35 ut and rotated them to 13 november .
this day was characterized by ongoing activity on the east limb .
we have built a composed off - limb euv image from the ratios of several eit 195 images observed from 00:00 ut to 05:00 ut , bypassing the intervals of activity .
the on - disk part is an average of six eit 195 images observed from 00:00 ut to 01:13 ut , in which the solar rotation was compensated .
the result is presented in figure [ f - f_lines_limb ] .
the light - blue lines correspond to the open field . in spite of the five - day difference and active conditions
, there is an overall correspondence between the calculated field lines and the off - limb structures . in summary , figures [ f - f_lines_center ] and [ f - f_lines_limb ]
demonstrate that the extrapolation in the potential approximation presents a more or less realistic picture .
furthermore , we observe that despite the ongoing activity , the distribution of coronal magnetic field remained relatively stable .
computed the maps of the magnetic helicity flux density injection and , in spite of the mixed helicity signs , showed the existence of a localized positive helicity injection in the southern part of ar 501 .
they also concluded that the positive helicity was ejected from this portion of the ar leading to the observed positive - helicity mc .
the accumulated positive helicity , as considered , was concentrated in one of the portions of the pre - eruptive filaments , while the helicity of other portions was negative and corresponded to the sign of the global helicity of ar 501 .
however , the idea of to account for the positive helicity of the near - earth mc does not seem to be convincing .
firstly , the helicity injection rate measured for some part of an active region during a rather short time interval does not provide information on the total helicity accumulated during the preceding evolution . secondly , the fresh idea of the authors about the presence of the opposite - helicity portions in the body of the pre - eruptive filament , in our opinion , does not have a convincing visual support .
finally , pointed out that the injection rate of the positive helicity flux was not high , so that such an injection was able to accumulate the helicity appropriate for the mc in , at least , six days . used the orientation of the filament barbs as a morphological indication of a section with a positive twist in the pre - eruptive filament ( figures [ f - shape_transform]a and [ f - shape_transform]b ) .
these authors found a filament segment consisting of two sections , which had the same direction of the central axis , but opposite twists .
the authors indicated , however , that this determination was relatively ambiguous . in our opinion , the identification of the filament barbs and their usage as a morphological indication is questionable for such a broad filament .
according to paper i , the filament was tilted by about @xmath40 to the solar surface , and features as the filament barbs could correspond to different parts located at different heights ( see figures 14 and 17 from paper i ) . to confirm the suggestions inferred from the apparent filament barbs , invoked different methods , _
i.e. _ , magnetic field extrapolation based on the linear ( constant @xmath10 ) force - free field approximation , and the computation of the magnetic helicity flux density injection .
using a magnetogram before the eruptions ( at 06:23 ut ) , they found a dominant negative @xmath10 by matching the shapes of the computed field lines to that of the observed coronal loops .
this finding confirms their conclusion that the large - scale magnetic field of ar 501 had a negative helicity sign .
the authors also compared the computed field lines with the post - flare arcade loops in the trace 171 full - resolution image @xcite at 09:43 ut and found that the loops could be modeled by using any small @xmath10 value positive , or negative , or null .
they also found that , despite the predominance in the ar 501 of negative @xmath10 values , corresponding to the negative global magnetic helicity , a small @xmath10 tended to be positive in the southeastern part of the ar .
the authors also computed an ongoing injection of the positive - helicity flux in this part of ar 501 and conjectured that this localized injection could be sufficient to make the magnetic helicity positive in this area , but its total value remained uncertain . modeled the post - flare loops visible in a half - resolution 195 image obtained with eit at 09:36 ut and found that the best value of @xmath10 was slightly positive along the total length of the visible arcade .
this led them to a conclusion that the ar had a global positive magnetic helicity .
the authors used the images obtained at the late post - flare stage , when the non - potentiality of the post - flare loops was least pronounced being , therefore , difficult to observe . on the other hand , at the earlier stages of a flare , more favorable in this respect , the loops are usually unresolved because of their small size .
we have succeeded in solving this problem .
the loops of a compact arcade visible in the trace images in figures [ f - helicity]a[f - helicity]c ( not analyzed by the aforementioned authors ) are skewed counter - clockwise in sunspot n3 and the region around it , indicating negative helicity .
the skewness of the loops in region s4 considered by seems to be the same , as expected for both ends of the same loops , connecting these regions in figures [ f - helicity]a[f - helicity]c .
the whole flare arcade has an inverse s - shape , which also corresponds to negative helicity .
g ) and positive ( continuous , @xmath41 g ) @xmath2 polarities in the bbso magnetogram used in the computations .
the top left gray box presents the ratio of the total negative to positive current helicity at the photospheric layer ( surf ) and in the coronal volume ( vol ) for the whole region .
the bottom right gray box is related to the framed region .
the axes are in arc seconds from the solar disk center . ] to verify this suggestion , we have computed the current helicity , @xmath42 , from a vector magnetogram of ar 501 observed at _ big bear solar observatory _ ( bbso ) on 18 november at 20:26 ut ( courtesy v. yurchyshyn ) .
the field of view of the magnetogram , of @xmath39 , was centered near the n1 sunspot ( about @xmath43 $ ] at 08:00 ut ) and covered the active region and its vicinity , but did not reach the region of bifurcation .
the line - of - sight component in the bbso magnetogram was considerably saturated in sunspots s1 ,
n3 , and s3 .
our computations were carried out in a spherical box with a photospheric base of @xmath44 and a height of 135 mm from both the raw magnetogram and a saturation - corrected one by using a temporally close soho / mdi magnetogram .
we used the method of the nlff extrapolation @xcite to compute two kinds of the current helicity maps .
the first kind of map , the photospheric @xmath45 density map , presents the distribution of @xmath45 at the photospheric layer .
the second kind of map is the column density distribution of @xmath45 .
each pixel of such a map is the photospheric base of a vertical column , whose height is 135 mm , and presents the total @xmath45 over all cells constituting this column .
comparison of the maps of the two kinds show that they are not very different , because the major contribution to the total current helicity is due to the lowest layers closest to the photosphere .
each attempt has resulted in a significant excess of the negative helicity in both the whole active region and in the framed area considered by , but with somewhat different quantitative results .
the saturation - corrected current helicity map is shown in figure [ f - helicity]d .
the ratios of the total negative to positive helicity for both the photospheric layer and for the total volume are specified in the gray boxes .
the top left box presents the ratios for the whole region , the bottom right box is related to the framed region .
indeed , the excess of the total negative helicity is less in the framed region , in agreement with the ongoing injection of the positive helicity found by , but still insufficient to make the total helicity in this region positive .
the absence of positive magnetic helicity in the localized area of ar 501 also indicates the absence of segments with the corresponding handedness in the body of the main filament f1 .
note also that the central part of the inverse s - shaped structure in figures [ f - helicity]a and [ f - helicity]b passes between sunspots n3 and s3 , where the negative current helicity is the largest in the whole ar 501 .
to make the situation clearer , we have additionally calculated magnetic field lines from the same nlff extrapolation , which was used in the computation of the current helicity .
it is difficult to reach a perfect correspondence between the field lines and the observed loops , because the rapid evolution of ar 501 during the 12.5 hours separating the euv image and the vector magnetogram could imply a change in the small - scale features , while the strong saturation of the vector magnetogram ( and , to a lesser degree , its limited field of view ) could considerably affect long loops .
for example , we have not succeeded to reproduce the long loops in the north
east part of the arcade ; however , its negative helicity is undoubted .
nevertheless , three sets of field lines with @xmath46 overlaid on the enlarged image of the arcade in figure [ f - helicity]b more or less correspond to the actually observed loops , in particular , the green lines overlapping the region of the questionable handedness .
all of these circumstances show that the attempts to reconcile the handedness of the magnetic cloud and active region 501 are not promising .
the question how the right - handed mc could be formed in the eruption of the left - handed filament remains unanswered .
the scenario proposed by corresponds to the concept of a simple eruption , when the internal magnetic helicity of a pre - eruptive structure , _
i.e. _ , its self - helicity ( or twist helicity ) , @xmath47 , determines the helicity sign of the interplanetary mc .
probably , this situation is typical of cmes associated with eruptions of quiescent filaments outside active regions .
however , if a filament ( flux rope ) erupts from an active region , then , under certain circumstances , the helicity of the mc can have a different origin .
this scenario implies the interaction and reconnection between magnetic fields of the eruptive structure with coronal magnetic fields surrounding the parent ar .
in such a case , a new magnetic structure is formed , which is the progenitor of the future interplanetary mc .
its helicity is determined by the sum of @xmath47 and the mutual ( or linkage ) helicity , @xmath48 , of the interacting magnetic fluxes .
_ via _ magnetic reconnection , the mutual helicity is transformed into the self - helicity of the mc .
depending on the sign of @xmath48 and the value of the @xmath49 ratio , the mc helicity can be different from the pre - eruptive structure not only in the value , but also in the sign .
such eruptions with mismatching helicity were suggested in the study by ; this scenario is also supported by the observations of the apparent disintegration of eruptive filaments mentioned in section [ s - introduction ] . to understand what could happen on the sun on 18 november , we first analyze the configuration of the magnetic field on spatial scales considerably larger than the size of the vector magnetogram in figure [ f - helicity ] .
this is possible by using full - disk magnetograms , of which line - of sight measurements ( _ e.g. _ , by soho / mdi ) are only available for 2003 .
thus , we had to use for this purpose mdi magnetograms and the potential field extrapolation .
the coronal field in a magnetic complex consisting of regions 501 , 503 , and their environment was computed in the potential approximation using the mentioned package @xcite for the full - disk soho / mdi magnetograms of 18 november observed at 06:23 , 07:59 , and 09:35 ut . the computations used 90 spherical harmonics .
the field line distribution computed from each of the magnetograms was similar ; we mainly use here the magnetogram at 09:35 ut .
figures [ f - field_lines][f - fil_motion ] present the coronal magnetic configuration . the s
n s n quadrupole in figures [ f - field_lines ] , [ f - top_view ] , and [ f - fil_motion ] is its basis .
figure [ f - top_view ] shows that the major eastern s - polarity sunspot has a large excess of negative magnetic flux which is unbalanced in this quadrupole .
a considerable portion of this flux is connected to remote sites on the solar surface ( figure [ f - fil_motion ] ) separated from ar 501 by the global polarity inversion line .
g , continuous @xmath50 g. the axes are in solar radii from the solar disk center . ]
a topological particularity of the coronal configuration is a magnetic null point located at a height of about 100 mm above the photosphere .
this is the only null point associated to large - scale magnetic fields on the visible side of the sun .
figure [ f - field_lines ] shows a side view of the complex of regions 501 , 503 , and their surroundings .
this complex was located at the solar disk center on 18 november .
figure [ f - field_lines]a shows the location of the null point ( slanted cross ) inside this complex , the pre - eruptive u - shaped filament f1 ( red semicircular arrow ) , and region of bifurcation rb , which had well - pronounced counterparts on the solar disk visible in the h@xmath10 line and in soft x - rays ( see paper i ) .
figure [ f - field_lines]b shows the magnetic field lines , which pass close to the null point .
according to the classification of , here we are dealing with a negative improper three - dimensional null point with the fan field lines rooted in the n polarities , and the field lines around the spine rooted in the s polarities .
the fan is perpendicular to the spine .
the region of bifurcation rb is rather close ( but not exactly co - spatial ) to the site , where the spine field line , which leaves the null point , enters the photosphere .
the lack of coincidence in our extrapolation , besides their possible actual difference , can be due to : a ) the usage of the potential approximation , and b ) the insufficient spatial resolution because of the limited number of the spherical harmonics .
there are indirect indications of a connection between rb and the null point : 1 ) rb firstly appeared as an isolated bright point in the h@xmath10 and sxi images one minute later than the u - shaped filament started to move at 07:41 ut .
the rb region and the filament were not connected by field lines .
their connection is possible _ via _ reconnection at the magnetic null point located between them .
invoking magnetic reconnection with a possible transformation of the null point into a current sheet is not the only possibility here .
the mhd disturbances generated by the initial displacement of the eruptive filament can cause local plasma heating in the vicinity of the null point by the accumulation and dissipation of the energy of fast - mode mhd waves or alfvn waves along the spine ( see , _ e.g. _ , the review by ; ) .
such heating augments the flux of heat and , possibly , that of non - thermal particles responsible for the increase of the emission intensity from the chromospheric and coronal plasmas above the photospheric base of the spine field line .
2 ) the rb region has appeared as a small ring with a brightening running clockwise ( paper i ) at the same time as the eruptive filament f1 would pass the magnetic null point ( this phenomenon started at 08:08 ut ) .
a similar situation with a considerably larger ring was observed in an eruptive event of 1 june 2002 @xcite . in that event ,
the magnetic configuration was different , it had a funnel - like shape , and the ring was situated above the photospheric footprint of a hemispheric separatrix surface with a magnetic null at its top .
the ` magnetic funnel ' confined all the ejections , as suggested by the movies of the event . similar to our case , a brightening running along the ring was observed in the 1 june 2002 event , when the eruptive filament , being transformed into a rotating ejection , passed through the magnetic null .
the motion of this ejection through the null must be accompanied by reconnection of the magnetic field lines associated to the ejection with those on the separatrix surface or nearby . a response to
this process was a concurrent euv disturbance , which propagated along the ring - like footprint of the separatrix surface . in the 18 november 2003 event
, the ring structure of region rb could be a manifestation of the photospheric base of a bundle of magnetic field lines twisting around , or close to the spine field line .
the twist indicates the presence of the longitudinal electric current along the spine and the inadequacy of the potential field approximation , which we use .
the twist is necessary for the brightening running along the ring that accompanies the concurrent magnetic reconnection of this bundle with an arbitrary magnetic rope or eruptive filament , for example , with a cylindrical shape .
a scheme in figure [ f - rotation ] illustrates the interaction .
as this figure shows , the bundle should be similar to a non - uniformly twisted cylindrical flux tube to produce a running brightening .
hoyle uniformly twisted tube with a non - constant @xmath10 does not satisfy this condition , while the lundquist magnetic cylinder with a constant @xmath10 satisfies it .
since rb is located within the zone of the negative magnetic polarity , the clockwise - running brightening corresponds to @xmath51 . in turn
, this indicates that the self - helicity of the magnetic flux around the spine is most likely negative . ,
@xmath52 , and @xmath53 are denoted with the colors presented in the top rectangle .
magnetic reconnection leads to a concurrent brightening running clockwise in the chromosphere and low corona above the rb region ( the brown round arrow at the bottom ) .
the axial magnetic fields , @xmath54 , of the eruptive flux rope and the bundle correspond to the positions of the thick blue and red loops at @xmath55 in figure [ f - scenario]a , respectively.,scaledwidth=60.0% ] 3 ) in the 1 june 2002 event @xcite , the interaction of the ejection with the magnetic ` funnel ' and its passage through the null appeared to have produced a wave disturbance in the corona observed as an ` eit wave ' rapidly expanding above the limb . in the 18 november 2003 event ,
the bifurcation of the eruptive filament was also accompanied by the appearance of a coronal wave , whose kinematical center was located high in the corona , above rb ( paper ii ) .
this was the second wave , not expected in the event .
the kinematical center of the first wave was within ar 501 and corresponded to the impulsive acceleration stage that is typical of impulsive eruptions @xcite .
until now , the contribution to the mc magnetic helicity from the relative position of the pre - eruptive filament and the structures outside of ar 501 have not been taken into account .
these structures got in contact at the magnetic null point , toward which the top of the expanding filament f1 moved .
figure [ f - field_lines ] illustrates the positions of the structures which could be involved in the eruptive process .
the values of the fluxes belonging to the west , east , north , and south magnetic dipoles are not known _ a priori_. note that each dipole represents only a fraction of the magnetic flux of a corresponding magnetic domain within the quadrupole configuration in figure [ f - field_lines ] .
it is also not known if the interaction between any of these dipoles and the eruptive filament was significant .
let us estimate by linking with which of these dipoles the pre - eruptive filament had the largest positive mutual helicity .
the presence of the positive mutual helicity is suggested by figure [ f - field_lines]b .
the curved blue arrow in the figure indicates the direction of the magnetic flux in the west dipole .
the red arrow indicates the direction of the axial magnetic field in the pre - eruptive filament .
let us invoke the right - handed screw rule and direct the screw axis along the blue arrow at its descending spine portion ; then the right - handed direction corresponds to the direction of the red axial field in the top of the filament .
if we put the screw axis along the top of the filament , then the result is the same ; the descending part of the blue arrow rotates clockwise .
therefore , one might expect that the west dipole and the pre - eruptive filament constitute a right - handed system . to estimate the mutual magnetic helicity , @xmath48
, we use the technique of interior angles @xcite . for this purpose
we need a top view of the magnetic configuration presented in figure [ f - field_lines]b .
such a view is shown in figure [ f - top_view ] , which represents the complex of regions 501 , 503 , and their environment located almost at the solar disk center when viewed from earth . the dotted lines @xmath56 ,
@xmath57 , @xmath58 , and the blue vector @xmath59 connect the positive and negative footpoints of the magnetic fluxes in the north , south , east , and west magnetic dipoles .
the red vector * f1 * connects the western ( negative ) and eastern ( positive ) ends of the erupted portion of the filament .
the sign of @xmath48 depends on the relation between the height @xmath60 of the pre - eruptive filament and those of the magnetic loops , @xmath61 , @xmath62 , @xmath63 , and @xmath64 , which constitute each of the dipoles .
the estimation shown below corresponds to the real case in which the height of the pre - eruptive filament is less than any of the other magnetic loops . for example , the mutual magnetic helicity of filament f1 and the west dipole @xmath59 is : @xmath65 where the angles @xmath66 are positive counter - clockwise , so that @xmath67 and @xmath68 .
the flux of the axial magnetic field in the pre - eruptive filament , @xmath69 , is unknown .
the magnetic flux of the west dipole , @xmath70 , will be estimated in section [ s - double_dimming ] . .
the red vector * f1 * connects the photospheric bases of the erupted portion of the filament in ar 501 .
the yellow dashed lines @xmath56 , @xmath57 , @xmath58 , and the blue vector @xmath71 connect the photospheric bases of the north , south , east , and west magnetic dipoles . @xmath72 and @xmath73 are the angular spans of vector @xmath59 measured from the ends of vector * f1*. the axes are in solar radii from the solar disk center .
the contour levels are symmetric relative to zero in steps of 35 g ; dotted @xmath74 g , continuous @xmath75 g. ] the mutual helicities @xmath76 can be estimated in a similar way . in particular , @xmath77 .
next , the angles between the lines @xmath78 , @xmath79 , and @xmath80 in figure [ f - top_view ] are small ; therefore , @xmath81 and @xmath82 . small changes of their mutual orientations can change the signs of the corresponding mutual helicities .
thus , @xmath83 appears to be the only solar source for a significant positive magnetic helicity of the interplanetary magnetic cloud .
if so , then the magnetic flux of the west dipole should be involved in the eruption of the u - shaped filament f1 .
this conjecture is confirmed observationally by the development of the double dimmings in the conjugate footpoints of the loops anchored in the west dipole ( see section [ s - double_dimming ] ) .
figure [ f - dimming_mag ] presents the development of dimming around 08:14 ut in association with the main eruption on 18 november .
the gray scale background in the panels shows three fixed - base ratios of eit 195 images , @xmath84 .
the contours outline the positive ( blue ) and negative ( red ) modeled flux concentrations of the coronal @xmath85 magnetogram . here
@xmath85 is the radial component of the coronal magnetic field computed at the spherical surface with a radius @xmath86 ( _ i.e. _ , at a height of @xmath87 mm above the photosphere ) from the soho / mdi magnetogram observed at 07:59 ut .
magnetogram extrapolated from an mdi magnetogram at a height of 28 mm ( blue n - polarity , red s - polarity ) .
the contour levels are @xmath88 g in panels ( a ) and ( b ) and @xmath89 g in panel ( c ) .
the yellow contour in panel ( a ) outlines the leading edge of the eruptive filament pouncing on a coronal obstacle above the bifurcation region rb denoted by the small oval .
the green contours in panel ( c ) outline the darkest dimming areas , which overlay a dipole outlined with a large oval .
the arrow shows the direction of the resulting magnetic field in the dipole . ]
figure [ f - dimming_mag]a presents the collision of the eruptive filament f1 ( yellow ) with a high coronal structure . their interaction is indicated by the brightenings turning around the bifurcation region ( rb ) underneath ( see paper i ) .
a large dimming south of ar 501 resulted from a preceding eruption and is beyond our scope .
figure [ f - dimming_mag]b ( 12 minutes later ) presents the outcome of the collision . both the brightenings and
region rb disappeared .
two dimming areas developed instead .
a new large dimming appeared at the previous position of the bifurcation region , and a weaker star - like dimming developed in ar 503 .
figure [ f - dimming_mag]c shows the situation observed still 12 minutes later .
the green contours outline the darkest portions of the two major west dimming regions .
the star - like dimming in ar 503 and the dimming encompassing the former bifurcation region overlaid a dipole pointing nearly south .
the magnetic flux in this dipole probably corresponds to the magnetic flux in a presumable eruptive structure .
the value of the magnetic flux computed within the dimming regions depends on the height , @xmath90 , to which the magnetogram is related .
this can be a photospheric magnetogram actually observed with mdi or a coronal @xmath85 magnetogram computed from the potential field extrapolation . to estimate the magnetic flux in the eruption
, we assume that the dipolar double dimming developed mainly due to the stretching of a closed magnetic flux tube , connecting these areas . one might expect that the positive and negative magnetic fluxes within the conjugate dimming areas are equal in absolute value .
the equality should be reached at a proper height above the photosphere . at the photospheric level , @xmath91
, the computation gives a negative imbalance of the magnetic fluxes , @xmath92 mx in the star - like dimming and @xmath93 mx in the large south dimming . while the height increases , the imbalance of the fluxes decreases .
it becomes zero at @xmath87 mm , where @xmath94 mx , and then becomes increasingly positive .
the height at which the flux is balanced is consistent with a typical height range of the coronal emission in the 195 line . considering each pole of the dipole as the centroid of the magnetic field distribution within each dimming region , we compute an orientation of the dipole of @xmath95 with respect to the north or about @xmath96 with respect to the ecliptic ( in the gse coordinate system ) . the gray arrow in figure
[ f - dimming_mag]c connects the centroids .
the direction of the arrow and its ends acceptably match the blue arrow of the west dipole in figure [ f - top_view ] .
our previous discussion provides a confirmation of the participation of the west dipole in the formation of the earth - directed mc with a positive helicity . the estimated magnetic flux , @xmath97 mx , is adequate to that in the magnetic cloud near earth , @xmath98 mx , estimated by .
the estimated magnetic flux in the west dipole is sufficient for the mc , while the two - fold excess could be lost by reconnection in the interplanetary space as proposed by and consistent with the fact that the mc crossed the sector boundary of the interplanetary magnetic field .
the inclination of the mc to the ecliptic plane estimated by different authors within a range of @xmath14 ( see ) corresponds to the orientation of the west dipole @xmath99 in figures [ f - top_view ] and [ f - dimming_mag]c .
according to our analysis in papers i , ii , and iv , the mc hitting the earth on 20 november had the following characteristics : i ) it was formed close to the solar disk center in the bifurcation of the eruptive filament , ii ) it was compact and disconnected from the sun , and iii ) it had an atypical spheromak - like configuration .
the outcome of section [ s - large_scale_config ] suggests that the mc could be formed by the interaction between , at least , two magnetic structures , whose mutual helicity was positive before the eruption .
these structures were the west dipole and the pre - eruptive filament .
the position of the west dipole on the solar disk meets requirement i ) .
an additional requirement follows from the total magnetic helicity conservation @xmath100 , where @xmath101 and @xmath102 are the self - helicities of the pre - eruptive filament and the west dipole , and @xmath103 is their mutual helicity .
the final configuration of the eruption depends on the sign of the total helicity . however , it is not possible to determine the sign of @xmath104 from the analysis of and our considerations in section [ s - helicity_absence ] .
the preceding estimates of the helicity sign were related to ar 501 only and did not contain information about the mutual helicity between the filament in the active region and magnetic structures outside of ar 501 .
the results of sections [ s - mutual_helicity ] and [ s - double_dimming ] show that if the magnetic fluxes of the west dipole , @xmath70 , and the filament , @xmath69 , are related as @xmath105 , then @xmath106 under typical assumptions .
thus , if the total helicity of two interacting fluxes is transformed through reconnection into the self - helicity of a single eruptive structure , then a right - handed mc can not be formed .
the situation is different , if the interaction and a chain of magnetic reconnections are followed by the formation of two eruptive structures rather than a single one . in this way ,
redistribution of the magnetic helicity is possible with almost the full transformation of the mutual helicity of two reconnected fluxes into the self - helicity of one of the resulting structures .
the second structure can carry away practically the whole negative helicity .
the mutual helicity of the two new eruptive structures approaches zero as they separate from each other , and therefore can be neglected . as a next step
we try to understand : i ) the possible geometry of the interaction between the eruptive filament f1 and the magnetic loops in the vicinity of the null point ( we associate its projection on the solar disk with the bifurcation region rb ) at their first contact and just after the passage of the null point , and ii ) why this interaction was followed by the visible dispersal over the solar surface of the cool plasma , which initially belonged to the eruptive filament f1 .
figure [ f - fil_motion ] shows a side view of the magnetic configuration , in which the main eruption occurred . for further considerations it is convenient to replace the erupting filament by a magnetic flux rope .
the rope has a toroidal ( axial ) and poloidal ( azimuthal ) components of the magnetic field .
the direction of motion of the middle part of the rope ( the solid red arrow marked 0 ) crosses the magnetic null point ( the green slanted cross ) .
the red dashed arrows marked 1 and 2 limit the cross section of the expanding rope . a selected magnetic field line ( thick blue )
denotes a magnetic loop of the west dipole , which was the major partner of the eruptive rope in the creation of the spheromak .
the thin contours on the solar surface correspond to the magnetic field values in steps of 34 g ; they are @xmath107 g ( dotted ) and @xmath108 g ( continuous ) away from the stronger - field regions near the limb .
the thick white circle denotes the solar limb .
the axes are in solar radii from the solar disk center . ]
the interaction between the eruptive magnetic flux rope and surrounding magnetic fields can result in two effects .
one is the redistribution of masses due to magnetic reconnection between the flux rope and the outer magnetic domain .
this domain contains field lines , which reach the solar surface far from the quadrupole . in the projection presented in figure [ f - fil_motion ] ,
the red solid line 0 is tangent to such field lines .
this effect is represented in figure [ f - reconnect_dispersal ] as implied in two - dimensional reconnection models .
another effect is a kinematic linkage between the interacting structures .
this essentially three - dimensional effect is used in a scheme presented in figure [ f - scenario ] . .
right : after magnetic reconnection .
the center of the rope displaces across the outer field @xmath109 , and cool plasma continues moving along @xmath109.,scaledwidth=80.0% ] figure [ f - reconnect_dispersal ] shows a cross section of the top of the magnetic flux rope .
this section moves along the external magnetic field @xmath110 with a velocity @xmath111 . in figure [ f - fil_motion ]
this situation corresponds to the position of the flux rope s center just after passing the magnetic null .
the two circles to the left of the figure are poloidal field lines of the flux rope .
cool plasma enclosed between them is denoted by the gray shading .
the situation after magnetic reconnection is shown to the right of the figure .
the center of the flux rope has shifted across @xmath110 , while plasma ( gray ) has departed from the flux rope and moves along the outer field @xmath110 .
the scheme in figure [ f - reconnect_dispersal ] is basically similar to figure 6 in .
the apparent difference between the schemes is due to different directions of motions of the eruptive filaments through the magnetic null point in the quadrupolar magnetic configuration . the 2d scheme in figure [ f - reconnect_dispersal ] does not change if we introduce an additional homogeneous magnetic field , @xmath112 , perpendicular to the plane of the figure without fixed ends and consider a non - compressive plasma . during magnetic reconnection ,
the frozen - in @xmath112 components of the total magnetic field are mixed like pens in a box without any effect on the process presented . however , this is not the case .
the magnetic field perpendicular to the plane of the figure is strongly inhomogeneous being concentrated in the curved toroidal flux rope , which continues its motion governed almost entirely by the toroidal propelling force .
reconnection creates new three - dimensional field lines between the footpoints of the eruptive flux rope and those of the outer field lines involved in the reconnection process .
the outcome is as follows : i ) the eruptive filament loses mass , ii ) the propelling toroidal force decreases faster than without reconnection , iii ) the disintegrating eruptive flux rope separates from cool plasma spreading out behind it .
we have found that the compact magnetic cloud in the 1820 november 2003 event could be the result of the redistribution of the magnetic helicity after the interaction between the left - handed eruptive filament and the west dipole @xmath113 .
the conclusion about the positive mutual helicity between the pre - eruptive filament f1 and the west dipole drawn in section [ s - large_scale_config ] implies the inequality @xmath114 .
this means that the height above the photosphere , @xmath60 , of the red loop , which represents filament f1 in figure [ f - field_lines ] , was less than that of the blue loop , which represents the loops belonging to dipole @xmath113 .
this was really the case before the eruption .
if the inequality would have been the reverse ; then , the sign of the mutual helicity would have changed . in an intermediate case ,
@xmath115 , the mutual helicity is zero .
the conservation of the total magnetic helicity restricts the choice of possible options for the subsequent dynamic reconfiguration of eruptive magnetic structures . in the absence of a clear idea about the formation mechanism of a compact right - handed spheromak - like
configuration ( called henceforth the spheromak for brevity ) , we use the following heuristic consideration . let us consider a straight trajectory along which the centroid of the eruptive filament moves .
arrow 0 in figure [ f - fil_motion ] corresponds to this trajectory across the magnetic null point .
the passage of the centroid through the null point ( whose position is assumed to be fixed ) corresponds to the equality @xmath116 , where @xmath117 is the height of the centroid .
this equality means that the mutual helicity between the moving filament and the loops anchored to the west dipole , @xmath118 , becomes zero . to keep @xmath119 until the onset of the interaction between these structures
, the trajectory of the eruptive filament should be below the null point .
this could correspond to , _
e.g. _ , arrow 2 in figure [ f - fil_motion ] , if the condition @xmath120 , which was valid before the eruption , is also valid in the dynamic regime . in this case ,
the clockwise angle between arrows 0 and 2 corresponds to a positive mutual helicity .
similarly , arrow 1 in figure [ f - fil_motion ] corresponds to a trajectory above the null point .
some consequences of this geometry including magnetic reconnection and mass depletion are shown in the two - dimensional figure [ f - reconnect_dispersal ] . the counter - clockwise angle between arrows 0 and 1 corresponds to a negative mutual helicity between the eruptive filament and the loops of the west dipole .
figure [ f - scenario ] presents a hypothetical 3d scheme for the formation of a right - handed spheromak , if the trajectory is below the null point .
the relative position of the red and blue thick solid loops in figure [ f - scenario]a is the same as in figure [ f - field_lines]b , which shows the situation before the eruption ( @xmath121 ) .
the active red loop represents the left - handed eruptive filament during subsequent times @xmath122 .
the passive blue loop is anchored to the west dipole involved in the eruption .
the red and blue dotted lines represent the modified shapes of the red and blue loops just after the onset of the interaction .
the expanding red loop embraces the leg of the blue loop rooted in rb and takes a y - like shape ( _ cf .
_ figure [ f - shape_transform ] ) .
the stretching of the red and blue magnetic structures in figure [ f - scenario]b leads to the formation of secondary loops which are linked .
magnetic reconnection results in the detachment in figure [ f - scenario]c of the secondary magnetic loops from their parent loops . a closed system of the orthogonal red and blue magnetic rings is formed . the blue ring developed from the magnetic flux of the west dipole .
we neglect its self - helicity and represent it as a system of non - twisted thin blue rings instead of a single thick blue ring .
the magnetic pressure separates these rings ( not interconnected ) from each other , and they get distributed along the red ring .
remnants of its negative helicity annihilate in this way ( figure [ f - scenario]d ) . a right - handed spheromak is formed .
the direction of its motion does not necessarily coincide with the direction , in which the eruptive filament , which takes back its shape , moves ( the thick red arrow ) .
b and [ f - top_view ] for the case when the trajectory ( black dashed line ) of the eruptive filament ( active red loop ) passes below the magnetic null point ( arrow 2 in figure [ f - fil_motion ] ) .
the blue loop is anchored to the west dipole involved in the eruption .
rb is the bifurcation region shown in figures [ f - shape_transform ] , [ f - field_lines ] , and [ f - dimming_mag ] .
( a ) @xmath121 : denotes the time before the eruption ; the interaction of the loops starts at @xmath55 .
the dotted lines at @xmath52 show the loops just after the onset of their interaction .
the expanding red loop embraces the leg of the blue loop rooted in rb and takes the y - like shape ( _ cf .
_ figure [ f - shape_transform ] ) .
( b ) the stretching of the red and blue magnetic structures results in the formation of secondary loops which are linked .
( c ) magnetic reconnection detaches the secondary loops from the parent ones .
a closed system of red and blue magnetic rings is formed .
( d ) the blue rings separate from each other and distribute along the red ring thus destroying remnants of its left helicity . a right - handed spheromak develops .
the apparent intersections of the blue and red loops in panels ( a)(d ) are due to the projection effect and do not imply the possibility of reconnection .
figure [ f - suggestion ] roughly outlines the dynamic reconfiguration presented in panels ( a ) and ( b ) of this figure , but viewed from a different angle . ]
the outcome of the process presented in figure [ f - scenario ] is the formation of two different eruptive structures sharing their positive and negative magnetic helicity . the first structure is a modification of the parent left - handed eruptive filament f1 , which recovers after the bifurcation and moves in the southwestern sector of the solar disk ( figure [ f - shape_transform ] ) . the second structure is the right - handed spheromak , which was created due to the catastrophe of the filament observed as its bifurcation . formally assuming that the effects presented in figures [ f - reconnect_dispersal ] and [ f - scenario ] coexisted in the real eruptive process , we come to the following conclusion .
the bifurcation of the eruptive filament f1 was accompanied by the decrease of its poloidal flux and mass depletion . the loss of mass accounts for the absence of a conspicuous core in cme2 , which appeared in lasco images in the same southwestern sector , where the eruptive filament f1 moved . the y - like trace of the mass , which has not left the sun , carries information about the deformation of the magnetic flux rope in the vicinity of the magnetic null point .
if the magnetic cloud , which reached earth , would have been associated with cme2 , then its magnetic helicity should be negative rather than positive .
this association seemed to be plausible in the first studies of the 1820 november 2003 event , but was ruled out in our paper ii . in our considerations , we called the ` spheromak ' a spheromak - like force - free magnetic configuration . the development of a such a configuration should be accompanied by a transformation of the magnetic energy excess of the interacting pre - spheromak structures into the kinetic energy of chaotic or directed plasma motions .
it is possible that some part of this kinetic energy had gone into the formation of a wave , possibly a shock , propagating away from the formation site of the spheromak .
an indication of such a wave is presented in paper ii ( see also section [ s - null_and_bifurcation_region ] , last paragraph ) .
while constructing the scheme in figure [ f - scenario ] , we pursued to have as a result the compact size and positive helicity of the developing magnetic structure to match the mc , which actually hit earth .
the spheromak meets these requirements .
however , our hypothetical scheme does not guarantee an earthward direction of the spheromak .
there are additional indications supporting that the spheromak has actually reached earth .
the formation of the spheromak in figure [ f - scenario ] is due to the interaction of different - temperature plasma structures , _
i.e. _ , the cool eruptive filament and the magnetic flux of the west dipole frozen into the coronal plasma . probably , this circumstance determined the atypically high inhomogeneity of the temperature distribution in the 20 november mc mentioned in section [ s - introduction ] .
the axial magnetic field of the spheromak in figure [ f - scenario ] is formed from the magnetic flux of the west dipole .
therefore , the orientation of the axial field of the spheromak should be close to the direction of the vector @xmath59 in figure [ f - top_view ] . in turn
, the inclination of @xmath59 to the ecliptic plane is reasonably close to the inclination of the mc ( section [ s - double_dimming ] , last paragraph ) . to conclude this section
, we note that the scheme of the interaction between the red and blue loops in the case of an upper trajectory 1 in figure [ f - fil_motion ] also provides the possibility to form a spheromak , but with a negative helicity .
most likely , this option did not occur in this event , as supported by the downward displacement of the central part of the eruptive flux rope in figure [ f - reconnect_dispersal ] . in the magnetic configuration presented in figure [ f - fil_motion ]
this effect works in the same direction both before the passage of the null point by arrow 0 and after it .
thus , the lower trajectory appears to be preferential and the only possibly one in this particular event .
the scenario of the 18 november 2003 event does not correspond to the concept of a simple eruption directly from ar 501 , in which the twist helicity of an eruptive structure or active region determines the handedness of the interplanetary magnetic cloud .
the nlff extrapolation of ar 501 shows an excess of negative twist , which is opposite to the positive sign of twist in the mc . to solve this contradiction ,
we have used the positive mutual helicity between the pre - eruptive filament and the flux tubes of a magnetic domain of the large - scale quadrupole configuration .
the interaction of these magnetic fluxes presumably occurred as the eruptive filament passed in the neighborhood of the coronal magnetic null point .
the positive mutual helicity of these two fluxes changed through magnetic reconnections into the positive self - helicity of a spheromak - like structure , whose geometry and parameters correspond to the magnetic cloud , which reached earth . in paper iv , we analyze the interplanetary disturbance responsible for the 20 november superstorm and outline the overall scenario of the whole event .
we thank v. yurchyshyn , who kindly supplied the vector magnetogram of ar 501 observed at bbso , and m. temmer for the h@xmath10 data .
we thank the co - authors of our papers i , ii , and iv , who are not involved in this study .
we appreciate an anonymous reviewer for valuable remarks and comments .
we are grateful to the instrumental teams of the kanzelhhe solar observatory ; trace and coronas - f missions ; mdi and eit on soho ( esa and nasa ) for the data used here .
this study was supported by the russian foundation of basic research under grants 11 - 02 - 00757 , 11 - 02 - 01079 , 12 - 02 - 00008 , 12 - 02 - 92692 , and 12 - 02 - 00037 , and the ministry of education and science of russian federation , projects 8407 and 14.518.11.7047 .
grechnev , v.v . ,
uralov , a.m. , chertok , i.m . ,
slemzin , v.a . ,
filippov , b.p . ,
egorov , ya.i . ,
fainshtein , v.g . ,
afanasyev , a.n .
, prestage , n. , temmer , m. : 2014b , _ solar phys .
_ _ 289 _ , 1279 .
grechnev , v.v . ,
uralov , a.m. , slemzin , v.a . ,
chertok , i.m . ,
filippov , b.p . ,
rudenko , g.v . ,
temmer , m. : 2014c , _ solar phys.___289 _ _ , 289 .
grechnev , v.v . ,
uralov , a.m. , slemzin , v.a . ,
chertok , i.m . ,
kuzmenko , i.v . , shibasaki , k. : 2008 , _ solar phys .
_ * 253 * , 263 .
slemzin , v.a . ,
kuzin , s.v . ,
zhitnik , i.a . ,
delaboudiniere , j .-
p . , auchere , f. , zhukov , a.n . , van der linden , r. , bugaenko , o.i . , ignatev , a.p . ,
mitrofanov , a.v . ,
pertsov , a.a . ,
oparin , s.n . ,
stepanov , a.i . ,
afanasev , a.n .
: 2005 , _ solar sys .
res . _ * 39 * , 489 .
srivastava , n. , mathew , s.k . , louis , r.e . ,
wiegelmann , t. : 2009 , _ j. geophys .
_ * 114 * , a03107 .
yurchyshyn , v. , hu , q. , abramenko , v. : 2005 , _ space weather _ , * 3 * , s08c02 .
zhang , m. , b. c. low : 2003 , _ astrophys .
j. _ * 584 * , 479 .
zhitnik , i.a . ,
bougaenko , o.i . ,
delaboudinire , j .-
, ignatiev , a.p . , korneev , v.v . ,
krutov , v.v . ,
kuzin , s.v . ,
lisin , d.v . , _
et al . _ : 2002 , _ proc .
10th european solar physics meeting , prague ( esa sp-506 ) _ , 915 . | our analysis in papers i and ii ( grechnev _ et al . _ , 2014 , solar phys .
289 , 289 and 1279 ) of the 18 november 2003 solar event responsible for the 20 november geomagnetic superstorm has revealed a complex chain of eruptions . in particular , the eruptive filament encountered a topological discontinuity located near the solar disk center at a height of about 100 mm , bifurcated , and transformed into a large cloud , which did not leave the sun .
concurrently , an additional cme presumably erupted close to the bifurcation region . the conjectures about the responsibility of this compact cme for the superstorm and its disconnection from the sun are confirmed in paper iv ( grechnev _ et al .
_ , solar phys . ,
submitted ) , which concludes about its probable spheromak - like structure .
the present paper confirms the presence of a magnetic null point near the bifurcation region and addresses the origin of the magnetic helicity of the interplanetary magnetic clouds and their connection to the sun .
we find that the orientation of a magnetic dipole constituted by dimmed regions with the opposite magnetic polarities away from the parent active region corresponded to the direction of the axial field in the magnetic cloud , while the pre - eruptive filament mismatched it . to combine all of the listed findings , we come to an intrinsically three - dimensional scheme , in which a spheromak - like eruption originates _ via _ the interaction of the initially unconnected magnetic fluxes of the eruptive filament and pre - existing ones in the corona . through a chain of magnetic reconnections
their positive mutual helicity was transformed into the self - helicity of the spheromak - like magnetic cloud . |
let @xmath2 be a smooth oriented spin 4-manifold with the rational homology of @xmath6 .
such manifolds play an important role in the study of homology cobordisms of homology 3-spheres and in addressing certain classification problems in 4-dimensional topology ; see discussion in @xcite .
their study , however , represents a challenge because the usual count of the seiberg
witten monopoles on @xmath2 generally depends on the auxiliary choices of metric and perturbation and hence does not result in a smooth invariant of @xmath2 .
this problem has been remedied by mrowka , ruberman , and saveliev @xcite , who defined an invariant @xmath0 for integral homology @xmath7 by adding an index - theoretic correction term to the count of the seiberg
witten monopoles on @xmath2 ; we will extend their definition to the case of rational homology @xmath7 in this paper . because of its gauge theoretic nature , the invariant @xmath0 is difficult to compute directly from its definition .
we address this problem in the current paper by expressing @xmath0 in terms of floer theoretic invariants via a gluing theory , the way it is done for the classical seiberg witten invariants .
an invariant relevant to this gluing theory was defined by fryshov @xcite : under the additional hypothesis that a generator of @xmath8 is carried by an embedded rational homology sphere @xmath9 , he showed ( * ? ? ?
* theorem 8) that the invariant @xmath10 arising from the monopole floer homology of @xmath9 with the induced spin structure @xmath11 is an invariant of the spin manifold @xmath2 alone ; we will denote this invariant by @xmath4 .
the invariants @xmath0 and @xmath4 are certainly different : for instance , the mod @xmath12 reduction of @xmath0 equals the rohlin invariant of @xmath2 , while this is not the case for @xmath4 .
the following theorem establishes a precise relation between @xmath0 and @xmath4 .
it is followed by some strong applications ( theorems [ t : psc ] , [ t : theta ] , and [ t : theta2 ] below ) to the study of metrics of positive scalar curvature and of the homology cobordism group of homology 3-spheres .
[ t : main ] let @xmath2 be a smooth oriented spin rational homology @xmath7 which is homology oriented by a choice of generator in @xmath13 .
assume that the poincar dual of this generator is realized by a rational homology sphere @xmath14 .
let @xmath11 be the induced spin structure on @xmath9 , and let @xmath15 be the spin cobordism from @xmath9 to itself obtained by cutting @xmath2 open along @xmath9 .
then @xmath16 the reduced monopole floer homology @xmath17 that appears in the statement of this theorem is the floer theory defined by kronheimer and mrowka @xcite , with rational coefficients .
we could use instead the monopole floer homology defined by fryshov @xcite , which would change the sign of the lefschetz number because of different grading conventions in the two theories .
the rational homology sphere @xmath9 in theorem [ t : main ] is oriented by the rule that the orientation on a curve hom - dual to @xmath18 followed by the orientation on @xmath9 gives the orientation on @xmath2 . the precise sign convention for @xmath0
is described in section [ s : lsw ] . in the special case of @xmath19 , where @xmath9 is an integral homology sphere , theorem [ t : main ] reduces to a theorem of fryshov (
* theorem 5 ) which relates the casson invariant @xmath20 , the fryshov invariant @xmath21 , and the euler characteristic of @xmath22 .
an analogous theorem ( * ? ? ? * theorem 7 ) holds for manifolds @xmath2 with @xmath23 once the term @xmath24 in formula ( [ e : main ] ) is replaced by the usual seiberg
witten invariant .
the proof of theorem [ t : main ] relies on the calculation of the two terms in the definition of @xmath0 , the count of the seiberg
witten monopoles on @xmath2 and the index - theoretic correction term , using metrics on @xmath2 with long necks @xmath25
. this neck stretching technique for counting monopoles is well known in gauge theory , although mainly in the separating case .
the non - separating case at hand was studied in ( * ? ? ?
* section 11.1 ) under the technical assumption of the absence of reducible monopoles on the non - compact manifold @xmath26 obtained by attaching infinite product ends to @xmath15 . in our case
, this assumption fails because @xmath27 .
instead , we prove that such a reducible monopole does not cause any trouble because it is isolated in the sense that a sequence of irreducible monopoles on @xmath2 can not converge to it as @xmath28 .
this is proved using an a priori estimate on the smallest eigenvalue of the dirac laplacian on a manifold with long neck .
we provide a detailed argument in sections [ s : eigenvalue ] , [ s : compact ] , and [ s : gluing ] , which all use the setup of the book @xcite . as a result , we are able to match the count of monopoles on @xmath2 with a certain lefschetz number in the monopole floer chain complex .
the truly novel part of the proof of theorem [ t : main ] , however , is the calculation of the correction term in @xmath0 , which boils down to calculating the index of the spin dirac operator on a manifold with periodic end modeled on the infinite cyclic cover of @xmath2 as @xmath28 .
this is done in two substantially different ways , one direct in section [ s : correction ] and the other using the end - periodic index theorem of @xcite in section [ s : eta ] .
a very useful technical result underpinning the proof of theorem [ t : main ] is the existence of riemannian metrics on the manifold @xmath26 with infinite product ends which make the @xmath29 sobolev completion of the spin dirac operator invertible .
this existence result , which is proved in section [ s : generic ] in all dimensions divisible by four , is an extension of the generic metric theorem of @xcite to certain non - compact manifolds .
one advantage of working with such generic metrics is that they greatly simplify the treatment of perturbations needed to ensure the regularity of the seiberg
witten moduli spaces , and allow to avoid perturbations on manifolds with periodic ends altogether .
see the discussion at the end of section [ s : lsw ] .
the splitting formula of theorem [ t : main ] makes the invariant @xmath0 computable in a number of cases .
this is due , on one hand , to the availability of advanced computational tools in monopole floer homology ( such as floer exact triangles and the @xmath30 symmetry ) , and on the other , to the identification between monopole floer homology and heegaard floer homology , by the work of kutluhan , lee , and taubes @xcite , or alternatively , the work of colin , ghiggini , and honda @xcite and taubes @xcite .
this newly found computability of @xmath0 leads to a number of applications , of which we present two in sections [ s : theorem b ] and [ s : theorem c ] . in both applications , the lefschetz number in formula vanishes , albeit for different reasons
the first application gives an obstruction to a @xmath1-manifold having a riemannian metric of positive scalar curvature .
historically @xcite , the seiberg
witten invariants have been used to produce many obstructions of this nature that go well beyond the classical index - theoretic obstruction of lichnerowicz @xcite .
we add to this body of knowledge the following theorem , which was originally proved by the first - named author ( * ? ? ?
* theorem 1.2 ) using different techniques .
it was conjectured in ( * ? ? ?
* remark 1 ) that there should exist a proof along the lines of this paper .
[ t : psc ] let @xmath2 be a smooth oriented spin rational homology @xmath3 which is homology oriented by a choice of generator in @xmath31 .
assume that the poincar dual of this generator is realized by a rational homology sphere @xmath14 with the induced spin structure @xmath11 .
then @xmath2 admits no riemannian metric of positive scalar curvature unless @xmath32 .
it was shown in @xcite that @xmath0 reduces modulo @xmath12 to the rohlin invariant @xmath33 .
as in @xcite , this fact leads to the corollary that , if @xmath2 admits a metric of positive scalar curvature , any rational homology sphere @xmath9 carrying the generator of @xmath8 must satisfy the relation @xmath34 with respect to the induces spin structure @xmath11 .
for example , if a generator of @xmath8 is carried by the brieskorn homology sphere @xmath35 , then @xmath2 can not admit a positive scalar curvature metric .
the second , and more elaborate , application is to the study of homology cobordisms .
recall that oriented 3-manifolds @xmath36 and @xmath37 are called homology cobordant ( respectively , @xmath38 homology cobordant ) if there exists a smooth , compact , oriented cobordism @xmath15 from @xmath36 to @xmath37 such that @xmath39 ( respectively , @xmath40 ) for @xmath41 .
the homology cobordism group @xmath42 is generated by oriented integral homology 3-spheres , modulo the equivalence relation of being homology cobordant .
similarly , the @xmath38 homology cobordism group @xmath43 is generated by oriented @xmath38 homology 3-spheres , modulo the equivalence relation of being @xmath38 homology cobordant .
let us first consider the group @xmath42 .
recall that the rohlin invariant provides a surjective homomorphism @xmath44 .
manolescu @xcite used @xmath30-equivariant seiberg
witten theory to show that this homomorphism does not split , that is , no integral homology sphere @xmath9 with @xmath45 has order two in @xmath46 .
it seems reasonable to conjecture that @xmath45 in fact implies that @xmath9 has infinite order in @xmath47 ; indeed , this was shown to be true for all seifert fibered homology spheres by the third - named author @xcite .
generalizing this result , we show that the conjecture holds under an additional assumption that @xmath9 is @xmath48-positive or @xmath48-negative : a @xmath38 homology sphere @xmath9 is said to be @xmath48-positive ( respectively , @xmath48-negative ) if the reduced monopole homology @xmath17 corresponding to the unique spin structure @xmath11 on @xmath9 is supported in degrees @xmath49 ( respectively , @xmath50 ) ; see definition [ defi : h - positive / h - negative ] for more details .
all seifert fibered homology spheres satisfy this assumption , and many more examples can be found in section [ s : theorem c ] .
[ t : theta ] any @xmath48-positive ( or @xmath48-negative ) integral homology sphere @xmath9 with @xmath51 has infinite order in @xmath47 .
it is worth mentioning that , instead of using theorem [ t : main ] , one could follow similar arguments to deduce theorem [ t : theta ] from stoffregen s connected seiberg witten floer homology @xcite .
let us now turn our attention to the group @xmath43 where we can prove a stronger result .
let @xmath52 be the subgroup of @xmath43 generated by all the @xmath53-spaces ( over @xmath54 ) that are @xmath38 homology spheres .
it was shown by stoffregen @xcite ( using the invariants @xmath18 , @xmath55 , and @xmath56 ) and hendricks , manolescu , and zemke ( * ? ? ?
* ; * ? ? ?
* proposition 1.4 ) ( using the invariants @xmath57 and @xmath58 ) that the brieskorn homology sphere @xmath35 has infinite order in the quotient group @xmath59 .
in particular , this implies that @xmath59 is an infinite group .
the following theorem generalizes this result by exhibiting a large family of such manifolds .
[ t : theta2 ] any @xmath48-positive ( or @xmath48-negative ) @xmath38 homology sphere @xmath9 with @xmath60 has infinite order in the group @xmath59 .
a further study of the @xmath48-positive ( @xmath48-negative ) condition using the rational surgery formula of ozsvth and szab @xcite leads to the following corollary .
[ c : surgery infinite order ] let @xmath9 be an integral homology sphere obtained by @xmath61 surgery on a knot @xmath62 in @xmath63 with @xmath64 odd , and suppose that @xmath65 . then @xmath9 has infinite order in @xmath46 in the following cases : 1 .
@xmath62 is the figure - eight knot ; 2 .
@xmath62 is a quasi - alternating knot with non - zero signature @xmath66 .
we remark that the fryshov invariant @xmath21 in the above corollary vanishes both in case ( 1 ) and in case ( 2 ) whenever @xmath67 ( see lemma [ property of thin knots ] ) hence the result of the corollary can not be proved using @xmath21 alone .
it may be possible , however , to give an alternative proof using other fryshov type invariants involving the @xmath30-symmetry , such as the invariants @xmath18 , @xmath55 , and @xmath56 of manolescu @xcite and the invariants @xmath57 and @xmath58 of hendricks and manolescu @xcite , once their behavior under the connected sum operation is better understood .
let @xmath9 be the integral homology sphere obtained by @xmath61 surgery on a two - bridge knot @xmath68 or @xmath69 with @xmath70 and odd @xmath71 .
then @xmath72 and @xmath9 has infinite order in the groups @xmath42 and @xmath59 .
another major reason to study the group @xmath43 is to gain information about the smooth knot concordance group @xmath73 via passing to the double branched cover of the knot .
recall that a knot @xmath62 is called khovanov - homology thin ( over @xmath74 ) if its reduced khovanov homology @xmath75 is supported in a single @xmath76-grading ( see @xcite ) .
such knots are very common : all quasi - alternating knots @xcite and 238 of the 250 prime knots with up to 10 crossings are khovanov - homology thin .
let @xmath77 be the subgroup of @xmath78 generated by the khovanov - homology thin knots .
theorem [ t : theta2 ] has the following curious corollary .
[ c : thin group infinite index ] any knot @xmath62 whose double branched cover satisfies the conditions of theorem [ t : theta2 ] generates an infinite cyclic subgroup in @xmath79 .
let @xmath62 be the @xmath80 or @xmath81 torus knot , @xmath82 .
then the connected sum @xmath83 ( @xmath84 ) is never smoothly concordant to any khovanov - homology thin knot . in our subsequent paper @xcite
, we will use similar techniques to compute the invariant @xmath0 for the mapping tori @xmath2 of all smooth orientation preserving involutions @xmath85 on integral homology spheres @xmath9 with the quotient @xmath63 .
that calculation will confirm the conjecture ( * ? ? ?
* conjecture b ) for all such mapping tori @xmath2 by identifying @xmath0 with its yang
mills counterpart , the invariant @xmath86 of furuta and ohta @xcite .
we begin by reviewing the definitions of the monopole floer homology and of the invariants @xmath0 and @xmath10 that go into the statement of theorem [ t : main ] in section [ s : prelim ] .
we also use this section to settle various technical matters and introduce some notations .
the proofs of theorems a , b , c and d , as well as corollaries [ c : surgery infinite order ] and [ c : thin group infinite index ] , are given in the three sections that follow .
they rely on certain technical results whose proofs are postponed until later in the paper for the sake of exposition .
the first of these results , discussed in section [ s : correction ] , is a calculation of the index - theoretic correction term @xmath87 on manifolds with long necks .
an alternative calculation using the end - periodic index theorem of @xcite is given in section [ s : eta ] .
both calculations boil down to computing the @xmath29 index of a spin dirac operator on certain manifolds with periodic ends , which we do in all dimensions @xmath88 .
similar techniques lead to the first eigenvalue estimates in section [ s : eigenvalue ] .
these are later used for the compactness and gluing results in sections [ s : compact ] and [ s : gluing ] .
several of these results rely on the generic metric theorem whose proof uses a rather different set of techniques and for this reason is postponed until section [ s : generic ] .
although many results in this paper can be generalized to @xmath89 structures , we will be mainly concerned with spin structures .
we will omit spin structures on 4-manifolds from our notations .
we will usually include spin structures on 3-manifolds , to be consistent with @xcite .
one exception are @xmath38 homology spheres : each of these has a unique spin structure which will be omitted .
* acknowledgments : * we are thankful to tye lidman , ciprian manolescu , and matthew stoffregen for generously sharing their expertise .
we begin by recalling the definitions of all the invariants involved and settling some technical matters .
let @xmath2 be an oriented smooth 4-manifold with rational homology of @xmath7 , equipped with a fixed spin structure .
we will review the definition of @xmath0 following @xcite .
note that the invariant @xmath0 was defined in @xcite only for an integral homology @xmath3 but a careful check of details shows that the construction of @xcite extends to a rational @xmath3 essentially word for word .
there are only two exceptions , which will be discussed in remark [ r : p - cyclic cover ] and remark [ r : dependence on spin str ] . with a slight abuse of language
, we will cite @xcite directly .
fix a homology orientation on @xmath2 by choosing a generator @xmath90 .
given a metric @xmath91 on @xmath2 and a co - closed 1-form @xmath92 orthogonal to @xmath93 ( the space of harmonic 1-forms on @xmath2 ) , consider the triples @xmath94 consisting of a @xmath95 connection @xmath96 on the determinant bundle of the spin bundle , a real number @xmath97 , and a positive spinor @xmath98 such that @xmath99 .
the gauge group @xmath100 acts freely on such triples by the rule @xmath101 . the blown - up seiberg
witten moduli space @xmath102 consists of the gauge equivalence classes of triples @xmath94 that solve the perturbed seiberg
witten equations @xmath103 the solutions are referred to as monopoles .
monopoles with @xmath104 are called reducible , and all other monopoles are called irreducible . the latter
are identified with the irreducible monopoles in the usual moduli seiberg
witten moduli space via the map @xmath105 .
according to ( * ? ? ?
* proposition 2.2 ) , the moduli space @xmath102 is regular for a generic choice of @xmath106 . in particular , @xmath102 is a compact oriented manifold of dimension zero that contains only irreducible solutions .
the count of points in @xmath102 with signs determined by the orientation and homology orientation is denoted by @xmath107 .
the invariant @xmath0 is defined in @xcite by the formula @xmath108 where @xmath109 is the correction term which cancels the dependence of @xmath107 on the parameters @xmath106 .
the definition of @xmath109 is as follows .
let @xmath14 be a connected manifold which is poincar dual to the chosen generator in @xmath13 .
note that @xmath9 is canonically oriented , and inherits a spin structure @xmath11 from @xmath2 .
denote by @xmath15 the cobordism from @xmath9 to itself obtained by cutting @xmath2 open along @xmath9 .
for any smooth compact spin manifold @xmath110 with spin boundary @xmath9 , consider the manifold @xmath111 with periodic end modeled on the infinite cyclic cover of @xmath2 ; each of the manifolds @xmath112 in this formula is just a copy of @xmath15 .
choose a metric and a perturbation on @xmath113 which match the metric and the perturbation over the end lifted from those on @xmath2 .
then the operator @xmath114 is fredholm with respect to the usual sobolev @xmath29 completion , and the correction term @xmath115 is independent of the choice of @xmath9 and @xmath110 .
[ r : p - cyclic cover ] the proof of well - definedness of @xmath109 in ( * ? ? ?
* proposition 3.2 ) makes use of the following fact : if @xmath2 is an integral homology @xmath7 , its @xmath116fold cyclic cover @xmath117 is a rational homology @xmath3 for any prime number @xmath116 .
this need not be true if @xmath2 is a rational homology @xmath7 .
however , one can use the argument of @xcite to show that @xmath117 is a rational homology @xmath3 as long as the prime number @xmath116 is large enough so that @xmath118 .
this is sufficient to complete the proof of well - definedness of @xmath109 .
theorem a of @xcite asserts that @xmath0 is independent of the choice of metric @xmath91 and generic perturbation @xmath55 , and that the reduction of @xmath0 modulo @xmath12 is the rohlin invariant of @xmath2 .
[ r : dependence on spin str ] unlike in the case of an integral homology @xmath3 treated in @xcite , different spin structures on @xmath2 may lead to different invariants @xmath0 . to keep our notations clean
, we will not include the spin structure in the notation .
note that when @xmath2 is a @xmath38 homology @xmath3 , different spin structures are all equivalent as @xmath89 structures and hence give the same invariant @xmath0 .
there are several implicit orientation conventions that go into the definition of @xmath0 .
we will not discuss them here but notice that altering these conventions consistently only changes @xmath0 by an overall sign .
the sign of @xmath0 was fixed in [ 32 , section 11.2 ] by the condition @xmath119 where @xmath120 is the casson invariant of an integral homology sphere @xmath9 normalized so that @xmath121 for the brieskorn homology sphere @xmath122 oriented as a link of complex surface singularity .
we conclude this section by addressing an important technical point about choices of metrics and perturbations . according to @xcite
, the operator @xmath123 can be made fredholm by choosing a generic metric @xmath91 and letting @xmath124 .
this choice of metric also guarantees ( * ? ? ?
* proposition 7.2 ) that the moduli space @xmath125 has no reducibles but not that it is regular .
one way to ensure regularity is to choose a generic perturbation @xmath55 small enough so that @xmath126
. then @xmath127 where @xmath87 stands for @xmath128 .
the perturbation @xmath55 in this formula can be replaced by a more general perturbation @xmath56 as in ( * ? ? ?
* section 9.2 ) or in section [ s : perturbation ] below .
as long as @xmath56 ensures regularity and is small enough , one can connect @xmath56 and @xmath55 by a generic path @xmath129 of small perturbations .
since the perturbations @xmath130 are small , the union @xmath131 contains no reducibles ( just like @xmath125 ) and provides an oriented cobordism between @xmath102 and @xmath132 , as explained in ( * ? ? ?
* proposition 9.4 ) .
this gives us the formula @xmath133 in this subsection , we will briefly recall the definition of the monopole floer homology @xmath134 .
we will focus on the special case when @xmath9 is a rational homology sphere and @xmath11 is a spin structure , which will suffice for the purpose of this paper
. the general definition can be found in kronheimer
mrowka ( * ? ? ?
* chapter 1 ) .
we will work with rational coefficients and omit the coefficient ring from our notations .
let @xmath9 be an oriented rational homology sphere with a riemannian metric @xmath48 and a spin structure @xmath11 .
an important example to have in mind is the rational homology sphere @xmath14 of theorem [ t : main ] with the induced spin structure @xmath11 .
trivialize the spinor bundle @xmath135 and choose the product connection @xmath136 to be our reference connection .
we will make the following assumption , which according to ( * ? ? ?
* theorem 1.1 ) holds for a generic metric @xmath48 .
[ generic metric on y ] the spin dirac operator @xmath137 has zero kernel .
let @xmath138 be the sobolev @xmath139 completion of the affine space of configurations @xmath140 , where @xmath141 is a connection in @xmath135 , @xmath142 is a spinor , and @xmath143 is an integer which will be fixed throughout the paper .
we refer to @xmath144 as the configuration space .
we also introduce the blown - up configuration space @xmath145 , which consists of the triples @xmath146 , where @xmath141 is a connection in @xmath135 , @xmath97 is a real number , and @xmath142 is a spinor with @xmath147 . let @xmath148 be the natural projection sending @xmath146 to @xmath149 . using the terminology of @xcite , configurations in @xmath144 are said to be downstairs , and those in @xmath145 are said to be upstairs .
a downstairs configuration @xmath140 is called irreducible if @xmath150 , while an upstairs configuration @xmath146 is called irreducible if @xmath151 .
all other configurations are called reducible .
the projection @xmath116 provides a diffeomorphism between the spaces of irreducible configurations upstairs and downstairs ; we denote both of these spaces by @xmath152 . the group of @xmath153 gauge transformations @xmath154 acts on @xmath144 by the formula @xmath155 and on @xmath145 by the formula @xmath156 .
both actions restrict to an action on @xmath152 .
the corresponding quotient spaces will be denoted by @xmath157 , @xmath158 , and @xmath159 .
note that @xmath159 is a hilbert manifold , while @xmath158 is a hilbert manifold with boundary .
the boundary of @xmath158 , given by the equation @xmath104 , contains the gauge equivalence classes of all reducible configurations @xmath160 $ ] .
the monopole floer homology was defined in @xcite as a variant of morse homology of the chern simons
dirac functional @xmath161 .
the definition of @xmath162 can be found in ( * ? ? ?
* definition 4.1.1 ) . to ensure that transversality holds
, @xmath162 is perturbed using a perturbation @xmath163 which is the formal gradient of a gauge invariant functional @xmath164 .
note that @xmath163 has two well - defined components , the connection component @xmath165 and the spinor component @xmath166 . the perturbed chern
dirac functional is denoted by @xmath167 .
its gradient @xmath168 gives rise to a vector field @xmath169 on @xmath170 .
let @xmath171 be any small number such that @xmath137 has no eigenvalues in the interval @xmath172 $ ] ( the existence of @xmath173 follows from assumption [ generic metric on y ] )
. then one can prove as in ( * ? ? ?
* proposition 2.8 ) that there exist perturbations @xmath163 satisfying the following assumption .
[ small perturbation ] the perturbation @xmath163 satisfies the following three conditions : a. @xmath163 is nice , that is , @xmath174 for all @xmath141 ; in other words , @xmath163 equals zero when restricted to reducible configurations , b. @xmath163 is admissible , that is , the critical points of @xmath169 are non - degenerate and the moduli spaces of trajectories connecting them are regular ; see definition 22.1.1 of @xcite , and c. the derivative of the spinor component @xmath175 of @xmath163 satisfies the inequality @xmath176 for any @xmath177 . under assumption [ small perturbation ] , the set @xmath178 of critical points of @xmath169 is discrete and can be decomposed into the disjoint union of three subsets : * @xmath179 : the set of irreducible critical points ; * @xmath180 : the set of reducible , boundary stable critical points ( i.e. , reducible critical points near which @xmath169 points inside the boundary ) ; * @xmath181 : the set of reducible , boundary unstable critical points ( i.e. , reducible critical points near which @xmath169 points outside the boundary ) . we will next take a closer look at the reducible critical points . according to ( * ? ? ?
* corollary 4.2.2 ) , the vector field @xmath182 has a unique reducible critical point @xmath183 $ ] downstairs , which we call @xmath184 $ ] .
the situation with the reducible critical points upstairs is quite different . to describe it ,
consider the perturbed dirac operator @xmath185 this is a self - adjoint elliptic operator . since @xmath163 is admissible ,
its eigenvalues are all non - zero and have multiplicity one .
we enumerate them so that @xmath186 for each @xmath187 pick an eigenvector @xmath188 of unit @xmath29-norm and let @xmath189 = [ ( b_0 , 0,\psi_i ) ] \in { \mathcal b}^{\sigma}(y)$ ] .
then @xmath190\mid i\geq 0\}\quad\text{and}\quad \mathfrak{c}^{u}=\{[\theta_{i}]\mid i<0\}.\ ] ] let @xmath191 ( respectively @xmath192 and @xmath193 ) be a vector space over @xmath194 with the basis @xmath195}\}$ ] indexed by the critical points @xmath196 $ ] in @xmath179 ( respectively @xmath180 and @xmath181 ) .
define a linear map @xmath197 by the formula @xmath198 } = \mathop{\sum}\limits_{[\beta]\in \mathfrak{c}^{o } } \#\breve{\mathcal{m}}([\alpha],[\beta])\cdot e_{[\beta ] } \ ] ] where @xmath199,[\beta])$ ] is the moduli space of unparameterized flow lines going from @xmath196 $ ] to @xmath200 $ ] and @xmath201,[\beta])$ ] is the signed count of points in this moduli space .
( this number is set to be zero if the dimension of the moduli space is positive . )
one defines maps @xmath202 and @xmath203 similarly .
consider the vector spaces @xmath204 the monopole floer homologies @xmath205 , @xmath206 and @xmath207 are defined , respectively , as the homology of the chain complexes @xmath208 , @xmath209 and @xmath210 with the differentials @xmath211 ( note that our formulas are simpler than those in @xcite because we are working with a rational homology sphere @xmath9 and a nice perturbation @xmath163 . ) the chain map @xmath212 with the matrix @xmath213 induces a natural map @xmath214 we define the reduced monopole floer homology as @xmath215 .
this is a finite dimensional vector space .
note that this definition matches the definition of the reduced monopole floer homology in ( * ? ? ?
* definition 3.6.3 ) because of the long exact sequence ( * ? ? ? * ( 3.4 ) ) . a rational homology sphere
@xmath9 is called an @xmath53-space ( over @xmath194 ) if the reduced monopole floer homology @xmath216 vanishes for all @xmath89 structures @xmath11 .
all these different versions of monopole floer homology are modules over the polynomial ring @xmath217 $ ] , where @xmath218 is a formal variable of degree @xmath219 .
in fact , we have canonical isomorphisms @xmath220 $ ] and @xmath221/u\cdot \mathbb{q}[u]$ ] as well as a non - canonical splitting @xmath222 .
in addition , the monopole floer homology has the so - called `` tqft property '' .
more precisely , any @xmath223 cobordism @xmath224 from @xmath225 to @xmath226 induces a morphism @xmath227 of @xmath217$]-modules , where @xmath228 stands for any one of the monopole floer homologies @xmath229 , @xmath230 , @xmath231 or @xmath232 .
each of these morphisms is induced by a respective chain map , whose definition requires further perturbations as described in section [ s : perturbation ] .
note that in the current paper , we only consider spin structures @xmath233 and omit them from our notations .
next , we need to discuss the canonical gradings in monopole floer homology . with each critical point
@xmath196\in \mathfrak{c}$ ] one associates two gradings , @xmath234)\,\in\,{\mathbb q}\quad \text{and } \quad { \operatorname{gr}}^{(2)}([\alpha])\,\in\,{\mathbb z}/2;\ ] ] see @xcite for the former and @xcite for the latter .
these naturally induce ( absolute ) @xmath54 and @xmath38 gradings on @xmath235 , @xmath236 and @xmath17 .
the generators @xmath189 \in \overline{hm}(y,{\mathfrak s})$ ] are graded by @xmath237)$ ] and @xmath238)$ ] if @xmath239 , and by @xmath237)-1 $ ] and @xmath238)-1 $ ] if @xmath240 . in all cases ,
the @xmath218-action decreases the @xmath54-grading by two and preserves the @xmath38 grading .
we will use the @xmath38 grading to define various lefschetz numbers , and use the @xmath54-grading to define the fryshov invariant .
the fryshov invariant @xmath10 is defined as negative one - half of the lowest @xmath194-grading of elements in @xmath241 , where @xmath242 is the map .
if @xmath2 is a spin homology @xmath7 as in theorem [ t : main ] , we define its fryshov invariant by the formula @xmath243 , where @xmath14 is an oriented rational homology sphere poincar dual to @xmath244 , with the induced spin structure @xmath11 .
it follows from ( * ? ? ?
* theorem 8) that @xmath4 is well - defined .
the fryshov invariant @xmath10 is an invariant of @xmath89 rational homology cobordism .
it is also known ( * ? ? ?
* theorem 3 ) that it changes sign with the change of orientation and that it is additive with respect to connected sums .
a heegaard floer version of @xmath4 was defined in @xcite without the assumption that @xmath9 be a rational homology sphere .
we will conclude this section by computing the gradings of the generators @xmath189 \in \overline{hm}(y,{\mathfrak s})$ ] . to this end , consider a smooth compact spin manifold @xmath110 with spin boundary @xmath9 and define @xmath245 where @xmath246 is the spin dirac operator on the manifold @xmath247 with cylindrical end obtained by setting @xmath19 in formula .
in essence , @xmath248 is a special case of the correction term and , like the latter , it is independent of the arbitrary choices in its definition .
[ l : reducible grading ] for any @xmath239 , we have @xmath238)=0 $ ] and @xmath237)=-2n(y , h,\mathfrak{s})+2i$ ] .
this follows directly from the definition of @xmath249 and @xmath250 ( see the discussion at the end of ( * ? ? ?
* page 421 ) ) .
the only non - trivial point is the use of assumption [ small perturbation](c ) to ensure that @xmath251 can be deformed into @xmath252 without acquiring non - zero spectral flow , which allows one to compute the grading with zero perturbation .
our proof will use the neck stretching operation which is well - known in gauge theory ; we will use its non - separating version . let @xmath2 be a spin rational homology @xmath7 and @xmath14 a rational homology sphere poincar dual to the choice of homology orientation @xmath253 .
the spin structure on @xmath2 induces a spin structure @xmath11 on @xmath9 .
let @xmath48 be a metric on @xmath9 satisfying assumption [ generic metric on y ] , and extend it to a metric @xmath91 on @xmath2 which takes the form @xmath254 in a product region @xmath255
\times y$ ] , @xmath256 .
given a real number @xmath257 , consider the spin manifold ` with long neck ' @xmath258 \times y)\ ] ] obtained by cutting @xmath2 open along @xmath259 and gluing in the cylinder @xmath260 \times y$ ] along the two copies of @xmath9 .
we also consider the non - compact manifold @xmath261\times y)\,\cup\,w\,\cup\,([0,+\infty)\times y)\ ] ] with two product ends .
the metric @xmath91 induces metrics on @xmath262 and @xmath263 , which will be denoted respectively by @xmath264 and @xmath265 .
[ a : dirac over w ] the metric @xmath91 on @xmath2 has the form @xmath254 in a product region @xmath255 \times y$ ] , @xmath256 , and makes the spin dirac operator @xmath266 invertible .
the existence of metrics satisfying assumption [ a : dirac over w ] will be proved in theorem [ t : two ] ; see also remark [ r : two ] .
note that , once assumption [ a : dirac over w ] is satisfied , the metric @xmath48 that shows up in its statement will automatically satisfy assumption [ generic metric on y ] .
the proof of theorem [ t : main ] will rely on the following two theorems about manifolds with long necks , whose proofs occupy section [ s : correction ] to section [ s : eta ] .
[ t : eta long neck ] let @xmath91 be a metric on @xmath2 satisfying assumption [ a : dirac over w ] . then , for all sufficiently large @xmath267 , the correction term @xmath268 in formula is well - defined , and we have the equality ( see ( [ e : correction term y ] ) ) @xmath269 [ t : sw = lefschetz ] let @xmath91 be a metric on @xmath2 satisfying assumption [ a : dirac over w ] .
then , for all sufficiently large @xmath267 and all sufficiently small perturbations @xmath270 ( defined in section [ s : perturbation ] ) which make the moduli space @xmath271 regular , we have the equality @xmath272 let @xmath91 be a metric on @xmath2 satisfying assumption [ a : dirac over w ] .
fix a large number @xmath273 and consider the truncated monopole chain complex @xmath274 for @xmath9 .
it is generated by the irreducible critical points and the boundary stable reducible critical points of grading @xmath275 .
then we have a short exact sequence @xmath276 where the chain complex @xmath277 is generated by the boundary stable reducible critical points of grading @xmath275 and has trivial differential . by lemma [ l : reducible grading ] , each @xmath278 with @xmath279 contributes a generator to @xmath277 .
therefore , we have @xmath280 the cobordism @xmath15 induces chain maps @xmath281 on the three chain complexes in the above exact sequence making the following diagram commute @xmath282 with the obvious abuse of notations , the lefschetz numbers of the three maps @xmath281 in this diagram are therefore related by the equation @xmath283 [ l : id ] the restriction of @xmath281 to the chain complex @xmath284 is the identity map .
this is essentially proved in ( * ? ? ?
* proposition 39.1.2 ) .
the result is stated there for homology but it holds as well for the chain complex because the boundary map is trivial for grading reasons . by lemma [ l : reducible grading ] ,
the @xmath38 gradings of the generators of @xmath284 are all zero .
then lemma [ l : id ] implies that @xmath285 and therefore @xmath286 on the other hand , for all sufficiently large @xmath287 , the group @xmath17 can be identified with the cokernel of the map @xmath288 therefore , we have the commutative diagram @xmath289 with the exact rows , which gives us the identity @xmath290 since @xmath291 is a finite length @xmath218-tail whose top grading is @xmath287 and lowest grading is @xmath292 , we have @xmath293 the restriction of @xmath281 on @xmath291 is the identity map by lemma [ l : id ] , therefore , @xmath294 and @xmath295 combining ( [ total lefs ] ) and ( [ lefs on u - tail ] ) with the fact that the lefschetz number of a chain map equals the lefschetz number of the induced map on homology , we obtain the identity @xmath296 and , after simplification , @xmath297 the proof is now complete because it follows from theorem [ t : eta long neck ] and theorem [ t : sw = lefschetz ] that , for all sufficiently large @xmath267 and the small perturbation @xmath298 , @xmath299 in this section we prove theorem [ t : psc ] , which is an application of theorem [ t : main ] to the question of existence of metrics of positive scalar curvature .
let @xmath2 be as in the statement of theorem [ t : psc ] and suppose that it admits a metric of positive scalar curvature . according to a theorem of schoen and yau @xcite ,
the poincar dual to the generator of @xmath300 can be realized by an embedded manifold @xmath301 which admits a metric of positive scalar curvature .
since the first chern class of the spin structure @xmath302 induced on @xmath303 vanishes , it follows from ( * ? ? ?
* proposition 36.1.3 ) that @xmath304 .
note that the manifold @xmath303 need not be a rational homology sphere , however , we will prove that its existence implies that @xmath305 .
our proof will adapt the argument of fryshov ( * ? ? ?
* section 13 ) that shows the well - definedness of the lefschetz number .
since @xmath303 generates @xmath8 , the standard covering space theory implies that the manifold @xmath303 lifts to the infinite cyclic cover @xmath306 , and that this lift can be arranged to be disjoint from a copy of @xmath9 .
it follows that , for some @xmath307 , the manifold @xmath308 contains a copy of @xmath303 separating its two boundary components .
therefore , the map @xmath309 factors through @xmath304 making @xmath281 nilpotent .
it then follows that the trace of @xmath281 vanishes in each @xmath38 grading , and that the lefschetz number of @xmath281 must therefore be zero .
we now prove theorems [ t : theta ] and [ t : theta2 ] from the introduction which assert that , in a number of circumstances , a homology sphere must have infinite order in the homology cobordism groups @xmath42 or @xmath59 .
the proofs can be found at the end of section [ s : obstruction ] .
the part of section [ s : theorem c ] after that is dedicated to examples and the proofs of corollaries [ c : surgery infinite order ] and [ c : thin group infinite index ] .
let @xmath310 be a rational homology sphere with a @xmath89 structure then @xmath17 is graded by the rational numbers , and we define the support of @xmath17 by the formula @xmath311 for the rest of this section , whenever we are considering the unique spin structure @xmath11 on a @xmath74 homology sphere @xmath9 , we will usually drop @xmath11 from our notations . in particular , the notations @xmath312 , @xmath313 and @xmath21 will be used , respectively , to denote the rohlin invariant , the support of reduced monopole floer homology , and the fryshov invariant for the unique spin structure .
[ cobordism obstruction ] let @xmath37 and @xmath314 be @xmath74 homology spheres such that @xmath315 and @xmath316 then @xmath37 is not @xmath74 homology cobordant to @xmath314 .
suppose to the contrary that we have an @xmath74 homology cobordism @xmath317 from @xmath37 to @xmath314 .
it carries a unique spin structure , which restricts to @xmath318 on @xmath319 . reversing the orientation , we obtain a spin cobordism @xmath320 from @xmath314 to @xmath37
now , consider the composite cobordism @xmath321 from @xmath37 to itself .
the morphism @xmath322 induced by the cobordism @xmath317 as in preserves the absolute grading . since the intersection @xmath323 is empty ,
we conclude that the map @xmath324 must be zero . by functoriality , the map @xmath325 is also zero ; in particular , its lefschetz number vanishes .
let @xmath2 be the homology @xmath3 obtained by identifying the two boundary components of @xmath15 via the identity map . then , using theorem [ t : main ] , we obtain @xmath326 this contradicts ( * ? ? ?
* theorem a ) which asserts that @xmath0 equals the rohlin invariant of @xmath2 modulo 2 . for a @xmath74 homology sphere
@xmath9 with spin structure @xmath11 , the condition @xmath327 is equivalent to condition that @xmath328 is odd .
the following definition is a slight generalization of the one given in the introduction .
[ defi : h - positive / h - negative ] let @xmath9 be a rational homology sphere with @xmath89-structure @xmath11 .
* we will say that @xmath310 is @xmath48-positive if @xmath17 is supported in degrees @xmath49 , that is , @xmath329 .
* we will say that @xmath310 is @xmath48-negative if @xmath17 is supported in degrees @xmath50 , that is , @xmath330 $ ] .
because heegaard floer homology is ( at present ) easier to compute , we would prefer to use it in place of the monopole floer homology in our calculations whenever possible .
in fact , these two theories are known to be isomorphic .
furthermore , by combining the main results of @xcite , the absolute @xmath54-gradings in the two theories coincide .
therefore , the relation @xmath331 between the fryshov invariant and the heegaard floer correction term holds for any rational homology spheres .
it then follows that the property of @xmath310 being @xmath48-positive ( respectively , @xmath48-negative ) can be characterized by saying that the reduced heegaard floer homology @xmath332 is supported in degrees @xmath333 ( respectively , in degrees @xmath334 ) .
[ seifert and figure eight ] by a positive orientation on a seifert fibered homology sphere we will mean the canonical orientation of it as a link of singularity . using the graded roots model for computing heegaard floer homology (
* section 11.13 ) , one can show that all seifert fibered homology spheres with positive orientation are @xmath48-negative , while the ones with negative orientation are @xmath48-positive . according to ( * ? ? ?
* proposition 8.3 ) , the homology sphere obtained by @xmath61 surgery on the figure - eight knot is @xmath48-negative if @xmath335 and @xmath48-positive if @xmath336 .
( note that @xmath53-spaces are both @xmath48-positive and @xmath48-negative . )
[ orientation reversal ] a @xmath74 homology sphere @xmath9 is @xmath48-positive if and only if the @xmath74 homology sphere @xmath337 obtained from @xmath9 by orientation reversal is @xmath48-negative .
we use @xmath11 and @xmath338 to denote the spin structure on @xmath9 and @xmath337 respectively . recall from ( * ? ? ?
* ( 3.4 ) ) that there is a long exact sequence and @xmath231 are consistent with those for @xmath339 and @xmath340 , respectively , the grading convention for @xmath230 differs from that for @xmath341 by @xmath342 .
for example , the generator of @xmath343 as a @xmath217$]-module has grading @xmath344 , while the generator of @xmath345 has grading @xmath219 .
we follow here the conventions of @xcite . ]
@xmath346 therefore , the set @xmath313 can be equivalently defined as @xmath347 . under the natural duality isomorphisms
@xmath348 the dual map @xmath349 is exactly the map @xmath350 , therefore , @xmath351 .
the result now follows because @xmath352 .
[ close under connected sum ] for @xmath74 homology spheres @xmath353 , the connected sum @xmath354 is @xmath48-positive ( respectively , @xmath48-negative ) if @xmath37 and @xmath355 are both @xmath48-positive ( respectively , @xmath48-negative ) .
because of lemma [ orientation reversal ] , we only need to treat the @xmath48-negative case .
let us introduce the notations @xmath356/u^{b})[-a]\quad\text{and}\quad \mathcal{t}^{-}_{a } = ( \mathbb{q}[u])[-a],\ ] ] where @xmath357 $ ] has degree 0 and for a graded module @xmath303 , we follow the convention @xmath358)_k = m_{k+c}$ ] for the grading shift .
let @xmath359 ( @xmath360 ) be the spin structure on @xmath361 . for both @xmath37 and @xmath314
, we have a ( non - canonical ) splitting of the @xmath362$]-modules , @xmath363 by combining the connected sum formula for heegaard floer homology ( * ? ? ?
* proposition 6.2 ) with the identification between monopole floer homology and heegaard floer homology ( or alternatively , by directly using the connected sum formula in @xcite ) we obtain @xmath364 } \widehat{hm}(y_2,{\mathfrak s}_{2}))[-1]\ , \oplus\ , ( { \operatorname{tor}}_{\,\mathbb{q}[u]}\,(\widehat{hm}(y_1,{\mathfrak s}_{1}),\widehat{hm}(y_2,{\mathfrak s}_{2}))[-2].\end{split}\ ] ] we will now trace the contributions of each of the summands of @xmath365 and @xmath366 to @xmath367 : * the tensor product @xmath368 } { \mathcal t}^{-}_{-2h(y_2)-1})[-1]\ , = \ , { \mathcal t}^{-}_{-2h(y_1)-2h(y_2)-1}\ ] ] contributes the infinite @xmath218-tail to @xmath369 ; * each of the tensor products @xmath370 } { \mathcal t}^{-}_{-2h(y_2)-1})[-1]\ , = \ , { \mathcal t}_{-2h(y_2)+a_i}(b_i),\\ & ( { \mathcal t}^{-}_{-2h(y_1)-1 } \otimes_{\,{\mathbb q}[u ] } { \mathcal t}_{c_j}(d_j))[-1]\ , = \ , { \mathcal t}_{-2h(y_1)+c_j}(d_j ) , \\ & ( { \mathcal t}_{a_i}(b_i ) \otimes_{\,{\mathbb q}[u ] } { \mathcal t}_{c_j}(d_j))[-1]\ ; = \ ; { \mathcal t}_{a_i+c_j+1}(\min(b_i , d_j))\end{aligned}\ ] ] contributes a summand to the kernel of the map @xmath371 ; * to compute @xmath372}{\mathcal t}_{c_{j}}(d_{j}))[-2]$ ] , consider the following _ grading preserving _ free resolution @xmath373 by taking tensor product with @xmath374 $ ] and computing homology of the resulting chain complex , we obtain @xmath375 } { \mathcal t}_{c_j}(d_j))[-2]\ , = \ , { \mathcal t}_{a_i + c_j + 2 - 2\max(b_i , d_j)}(\min(b_i , d_j)),\ ] ] which contributes another summand to the kernel of @xmath376 .
since @xmath37 and @xmath314 are both @xmath48-negative , we have @xmath377 and @xmath378 .
also note that @xmath379 , @xmath380 .
it is now easy to check that all the summands in @xmath381 are supported in degrees at most @xmath382 .
therefore , @xmath383 is @xmath48-negative , and the lemma is proved .
[ c : first part ] let @xmath37 be an @xmath48-positive ( respectively , @xmath48-negative ) @xmath74 homology sphere , and suppose that @xmath384 . then @xmath37 is not @xmath74 homology cobordant to any @xmath48-negative ( respectively , @xmath48-positive ) @xmath74 homology sphere @xmath314 .
suppose to the contrary that @xmath37 is @xmath74 homology cobordant to an @xmath48-negative @xmath74 homology sphere @xmath314 .
since both @xmath48 and @xmath385 are invariants of @xmath74 homology cobordism , @xmath386 and in particular @xmath387 .
let @xmath388 then @xmath389 and @xmath390 $ ] so that @xmath391 .
this contradicts proposition [ cobordism obstruction ] .
[ h - positive not cobordism to h - negative ] let @xmath392 be @xmath48-positive @xmath74 homology spheres , and @xmath393 positive integers .
suppose that at least one of the @xmath361 satisfies the condition @xmath394
. then the connected sum @xmath395 can not be @xmath74 homology cobordant to @xmath63 .
a similar result holds for @xmath48-negative homology spheres .
suppose to the contrary that the connected sum ( [ connected sum manifold ] ) is @xmath74 homology cobordant to @xmath63 .
without loss of generality , we may assume that @xmath396 .
then the manifold @xmath397 , which is @xmath48-negative by lemma [ orientation reversal ] , is @xmath74 homology cobordant to the manifold @xmath398 which is @xmath48-positive by lemma [ close under connected sum ] .
this contradicts corollary [ c : first part ] .
suppose to the contrary that @xmath9 is of finite order in @xmath47
. then @xmath399 .
this contradicts corollary [ h - positive not cobordism to h - negative ] since a @xmath400 homology cobordism is also a @xmath74 homology cobordism .
suppose to the contrary that @xmath9 is of finite order in @xmath59 .
then there exists an integer @xmath82 and an @xmath53-space @xmath401 which is a @xmath74 homology sphere , such that @xmath402 is @xmath74 homology cobordant to @xmath403 .
this contradicts corollary [ h - positive not cobordism to h - negative ] since @xmath401 is both @xmath48-positive and @xmath48-negative . using the spectral sequences of ozsvth - szab @xcite and bloom
@xcite one can easily see that , for any khovanov - homology thin knot @xmath404 , the double branched cover @xmath405 is an @xmath53-space over @xmath74 .
the universal coefficient theorem then implies that @xmath405 is also an @xmath53-space over @xmath194 .
note that the double branched cover of @xmath406 with branch set a smooth concordance between two knots is a @xmath38 homology cobordism between the double branched covers of the knots .
the result now follows from theorem [ t : theta2 ] . in this section
, we will use the rational surgery formula of ozsvth and szab @xcite to obtain a sufficient condition for a surgered manifold to be @xmath48-positive .
corollary [ c : surgery infinite order ] will be proved in the next section by checking this condition and applying theorem [ t : theta ] .
let @xmath62 be a knot in @xmath63 .
given co - prime integers @xmath116 and @xmath407 , denote by @xmath408 the manifold obtained by the @xmath409 surgery on @xmath62 .
for any @xmath410 , the manifold @xmath411 is a rational homology sphere .
it admits exactly @xmath116 distinct @xmath223 structures , which can be naturally identified @xcite with the elements of @xmath412 . the @xmath223 structure on @xmath411 corresponding to an integer @xmath413 will be denoted by @xmath414 $ ] .
[ stable heegaard floer ] for all sufficiently large @xmath410 and all @xmath415 with @xmath416 the heegaard floer homology group @xmath417)$ ] , viewed as a relatively graded @xmath217$]-module , is independent of @xmath116 .
[ surgery h - positive ] suppose that , for all sufficiently large @xmath410 and all @xmath415 with @xmath416 , the rational homology sphere @xmath418)$ ] is @xmath48-positive .
then , for any positive integer @xmath64 , the integral homology sphere @xmath419 is @xmath48-positive .
this is a straightforward corollary of ozsvth - szab s rational surgery formula @xcite . for the sake of completeness
, we will sketch the argument here and refer the reader to @xcite for a concise summary .
( see also @xcite , which treats a similar situation as here ) . for a sufficiently large @xmath410 and any @xmath420 ,
consider @xmath64 copies @xmath421 of the heegaard floer homology @xmath422)$ ] ( this notation is justified by theorem [ stable heegaard floer ] ) , and @xmath64 copies @xmath423 of the module @xmath424\big/(u\cdot { \mathbb q}[u])$ ] . by (
* theorem 1.1 ) and ( * ? ? ?
* remark 2.3 ) , one can recover the heegaard floer homology @xmath425 as the homology of the mapping cone of a certain map @xmath426 in practice , one can take a large enough integer @xmath287 and instead consider the mapping cone of the truncated map @xmath427 this map is surjective so one has an isomorphism @xmath428 furthermore , one can impose suitable absolute gradings on @xmath429 such that the above isomorphism preserves the absolute grading . recall that @xmath429 admits a splitting @xmath430 .
let @xmath431 be the absolute grading of the bottom term in @xmath432
. then @xmath433 by ( * ? ? ?
* ( 2.1)@xmath434(2.4 ) ) .
using the assumption that the @xmath223 structure @xmath414 $ ] is @xmath48-positive for all sufficiently large @xmath410 , we conclude that @xmath429 is supported in degrees @xmath435 .
this implies that @xmath436 is supported in degrees greater than or equal to @xmath437 therefore , the integral homology sphere @xmath419 is @xmath48-positive . in this section
, we will apply proposition [ surgery h - positive ] to the classes of @xmath53-space knots and floer homology thin knots , and then prove corollary [ c : surgery infinite order ] .
recall that a knot @xmath62 is called an @xmath53-space knot ( over the rationals ) if there is a rational number @xmath438 such that the manifold @xmath408 is an @xmath53-space , that is , @xmath439 this condition actually implies that @xmath440 is an @xmath53-space for all @xmath441 .
proposition [ surgery h - positive ] has the following corollary .
[ surgery on l - space knot h - positive ] let @xmath62 be an @xmath53-space knot .
then @xmath419 is @xmath48-positive for all @xmath82 .
using corollary [ surgery on l - space knot h - positive ] , one can derive a result similar to corollary [ c : surgery infinite order ] for all @xmath53-space knots .
however , this can be proved directly using the heegaard floer correction term .
now we turn to the case of surgeries on floer homology thin knots .
we will need a number of constructions involving heegaard floer homology of knots , for which we refer to the original paper @xcite , as well as to the survey @xcite .
recall that , for an even integer @xmath442 , a knot @xmath62 is called floer homology @xmath442-thin ( over the rationals ) if the bigraded knot floer homology group @xmath443 satisfies the condition @xmath444 the following lemma summarizes properties of thin knots that are useful for the application we have in mind .
[ property of thin knots ] let @xmath442 be an even integer , @xmath62 a floer homology @xmath442-thin knot , and @xmath64 a positive integer
. then 1 .
in particular , @xmath446 if and only if @xmath447 ; 2 . if @xmath448 then for all sufficiently large @xmath410 and all integers @xmath413 such that @xmath449 the rational homology sphere @xmath418)$ ] is @xmath48-positive .
recall that the knot floer complexes are generated by triples @xmath450 $ ] satisfying various conditions , where @xmath242 and @xmath451 are integers , and @xmath452 is an intersection point between lagrangian tori in the symmetric product of the heegaard surface . for any @xmath453
, we will denote by @xmath454 the complex generated by triples @xmath450 $ ] with @xmath455 , @xmath456 .
we will use similar notations for the other complexes .
it follows from ( [ thin condition ] ) that @xmath457 is supported in degree ( absolute maslov grading ) @xmath458 . via a basic spectral sequence argument
, this implies that @xmath459 with these two facts established , we can prove that @xmath460 by repeating word for word the proof of ( * ? ? ?
* corollary 1.5 ) ( which deals with the special case of an alternating knot @xmath62 ) . since @xmath461 for any @xmath82 ( see ( * ? ? ?
* proposition 1.6 ) ) , claim ( 1 ) is proved .
we now turn to claim ( 2 ) .
since the spin@xmath462 structures @xmath414 $ ] and @xmath463 $ ] are conjugate to each other , one has an isomorphism @xmath464 ) \cong hf^{+}(s^3_p ( k),[-s])$ ] .
therefore , it is sufficient to consider the case of @xmath465 . recall from @xcite that there is an isomorphism @xmath466)\ , \cong\ , h(c\{\max(i , j - s)\geq 0\}).\ ] ] of relatively graded @xmath217$]-modules .
for any integer @xmath467 , denote by @xmath468 the graded module @xmath469/u\cdot \mathbb{q}[u])[-a]$ ] ( cf .
( [ u - tail ] ) )
. then we have a decomposition of absolutely graded @xmath217$]-modules , @xmath470 for some integer @xmath471 and a finite dimensional @xmath194-vector space @xmath472 .
consider the short exact sequence @xmath473 since @xmath474 is supported in degrees @xmath475 , we obtain @xmath476 with the last equality following from the isomorphisms @xmath477\cong { \mathcal t}^{+}_{2s}.\ ] ] therefore , @xmath478 is supported in degrees @xmath479 .
the proof will be complete once we show that @xmath480 .
to this end , consider another short exact sequence @xmath481 since @xmath482 is finite dimensional , for any sufficiently large integer @xmath483 we have isomorphisms @xmath484 let @xmath485 be any non - zero element .
since @xmath486 , we have @xmath487 .
this implies that @xmath488 and @xmath489 because @xmath478 is supported in degrees @xmath479 we conclude that @xmath480 , which completes the proof .
[ surgery on thin knot is h - positive ] let @xmath62 be a floer homology @xmath442-thin knot with @xmath490
. then @xmath491 is @xmath48-positive for all @xmath71 .
this is immediate from lemma [ property of thin knots ] ( 2 ) and proposition [ surgery h - positive ] .
since @xmath492 and @xmath64 is odd , it follows from the surgery formula for the rohlin invariant @xcite that @xmath493 .
claim ( 1 ) now follows from example [ seifert and figure eight ] . to prove claim ( 2 ) , consider the mirror image @xmath494 of the knot @xmath62 .
since @xmath494 is quasi - alternating and @xmath495 , it is sufficient to consider the case of @xmath71 .
according to @xcite , any quasi - alternating knot @xmath62 is floer homology @xmath66-thin over @xmath38 , where @xmath496 stands for the knot signature . by the universal coefficient theorem
, this implies that @xmath62 is also floer homology @xmath66-thin over @xmath54 .
if @xmath497 , it follows from lemma [ property of thin knots ] ( 1 ) that @xmath498 and hence @xmath491 has infinite order in @xmath47 . if @xmath499 , it follows from corollary [ surgery on thin knot is h - positive ] that @xmath491 is @xmath48-positive and hence it has infinite order in @xmath47 .
in this section , we will prove theorem [ t : eta long neck ] .
the index theory that will go into our proof is not specific to dimension four , therefore , we will work in more generality than strictly necessary .
let @xmath2 be a connected smooth spin compact manifold of dimension @xmath500 with a primitive cohomology class @xmath501 .
let @xmath14 be a connected manifold poincar dual to @xmath56 with the induced spin structure @xmath11 .
choose a metric @xmath91 on @xmath2 which takes the form @xmath254 in a product region @xmath255
\times y$ ] , @xmath256 .
we will assume that @xmath502 is a spin boundary and that the @xmath503genus of @xmath2 vanishes ; both of these conditions are automatic when @xmath2 is a homology @xmath7 . given a real number @xmath257 , construct the spin manifold @xmath504
\times y)\ ] ] as in by cutting @xmath2 open along @xmath259 and gluing in the cylinder @xmath260 \times y$ ] along the two copies of @xmath9 .
the metric @xmath91 defines a metric @xmath505 on @xmath506 , which lifts to a metric @xmath505 on the infinite cyclic cover of @xmath506 determined by @xmath56 .
following , denote by @xmath507 the manifold with _ periodic _ end modeled on this infinite cyclic cover , and by @xmath247 and @xmath26 the manifolds with _ product _ ends modeled on the product @xmath508 with metric @xmath509 .
note that @xmath26 has two ends , corresponding to the two boundary components of @xmath15 .
the metrics will often be suppressed in our notations .
[ thm ] assume that the spin dirac operator @xmath510 is an isomorphism .
then for all sufficiently large @xmath267 the end - periodic operator @xmath511 is fredholm of index @xmath512 the existence of metrics on @xmath26 making the operator @xmath513 invertible is addressed in theorem [ t : two ] . when applied to a spin @xmath1-manifold @xmath2 with the rational homology of @xmath3 , theorem [ t :
eta long neck ] is a straightforward corollary of theorem [ thm ] .
we begin by proving two technical lemmas which will be used later in the argument .
[ norm control ] suppose @xmath514 is a surjective bounded operator between hilbert spaces
. then there exists a constant @xmath515 such that for any vector @xmath516 one can find a vector @xmath517 with @xmath518 . by the open mapping theorem
the map @xmath519 is an isomorphism .
[ kernel to domain ] let @xmath520 and @xmath521 be bounded linear operators between hilbert spaces , and assume that @xmath522 is surjective
. then the operator @xmath523 is fredholm if and only if the operator @xmath524 is fredholm and @xmath525 the projection map @xmath526 can be included in the short exact sequence @xmath527 , which is naturally a subsequence of the short exact sequence @xmath528 .
the quotient sequence @xmath529 is exact by the snake lemma , which proves the equality of the cokernels of the two operators in question .
the equality of their kernels is clear
. we will find it convenient to introduce the notation @xmath530\,\times\,y$ ] and write @xmath531 and @xmath532 with @xmath533 and @xmath534 for all @xmath535 .
each of the manifolds @xmath536 is just a copy of @xmath9 but the notations are chosen so that @xmath537 is a cobordism from @xmath538 to @xmath539 while @xmath540 is a cobordism from @xmath539 to @xmath541 .
the spin dirac operator @xmath542 is a self - adjoint elliptic operator on a compact manifold hence it has a discrete spectrum with real eigenvalues of finite multiplicity .
denote by @xmath543 the subspaces spanned by the eigenspinors of @xmath542 with respectively the positive and the negative eigenvalues .
the @xmath29 orthogonal projections onto these subspaces will be denoted by @xmath544 .
[ assum ] the operator @xmath513 of theorem [ thm ] is invertible if and only if the following two conditions are satisfied : 1 .
the dirac operator @xmath542 has zero kernel , and 2 .
the dirac operator @xmath545 with the atiyah patodi singer boundary conditions is an isomorphism .
here , @xmath546 denote the restriction maps to the boundary components @xmath547 and @xmath548 of @xmath15 .
the condition on @xmath542 to have zero kernel is equivalent to the condition on @xmath513 to be fredholm .
the relation between @xmath513 and the operator @xmath549 with the atiyah patodi
singer boundary conditions is well known ; see ( * ? ? ?
* proposition 3.11 ) . from now on
, we will assume that the operator @xmath513 is invertible or , equivalently , that the conditions ( 1 ) and ( 2 ) of lemma [ assum ] are satisfied .
given a family of hilbert spaces @xmath550 , @xmath551 , their direct sum @xmath552 is the hilbert space which consists of all the sequences @xmath553 of vectors @xmath554 such that @xmath555 , the inner product of sequences @xmath553 and @xmath556 being @xmath557 .
any uniformly bounded family of bounded operators @xmath558 gives rise to a well defined bounded operator @xmath559 of norm @xmath560 .
an application of this abstract construction to the above splitting of @xmath507 yields the following result ( we suppress spinor bundles in our notations ) .
[ break the sobolev space ] the natural restriction maps provide hilbert space isomorphisms @xmath561 where @xmath562 is the restriction map @xmath563 which sends @xmath564 to @xmath565 claim ( 1 ) is straightforward . to prove ( 2 ) , observe that there is an obvious norm preserving inclusion of @xmath566 into @xmath567 .
the result now follows from the fact that all spinors in @xmath567 belong to @xmath568 , see for instance manolescu ( * ? ? ?
* lemma 3 ) .
[ surjective ] ( 1 ) the dirac operator @xmath569 is surjective .
\(2 ) the operator @xmath570 sending @xmath142 to @xmath571 is an isomorphism .
\(3 ) the restriction maps @xmath572 of lemma [ assum ] are surjective .
\(4 ) the restriction map @xmath562 of lemma [ break the sobolev space ] is surjective .
\(5 ) the operator @xmath573 is surjective .
claim ( 1 ) is proved in ( * ? ? ?
* corollary 17.1.5 ) ; ( 2 ) can be easily verified using the spectral decomposition of @xmath542 and the fact that @xmath574 ; ( 3 ) is a standard fact about sobolev spaces ; ( 4 ) follows from ( 3 ) and lemma [ norm control ] ; ( 5 ) follows from ( 1 ) and lemma [ norm control ] .
the proof will essentially be a repeated application of lemma [ kernel to domain ] to the dirac operator @xmath575 * step 1 .
* consider the operator @xmath576 sending @xmath577 to @xmath578 it follows from lemma [ break the sobolev space ] , lemma [ surjective ] ( 4 ) and lemma [ kernel to domain ] that @xmath579 is fredholm if and only if @xmath580 is fredholm , and @xmath581 * step 2 . *
observe that the kernel of @xmath582 equals @xmath583 and consider the operator @xmath584 sending @xmath577 to @xmath585 it follows from lemma [ kernel to domain ] and lemma [ surjective ] ( 5 ) that the operator @xmath580 is fredholm if and only if @xmath586 is fredholm , and @xmath587 * step 3 .
* using the subspaces @xmath588 and @xmath589 spanned by the positive and negative eigenspinors of the operator @xmath542 , and the respective @xmath29 orthogonal projections @xmath590 and @xmath591 , the operator @xmath586 can be written as the operator @xmath592 sending @xmath577 to @xmath593 since the operators @xmath586 and @xmath594 are isomorphic , we have @xmath595 * step 4 .
* by lemma [ surjective ] ( 2 ) , for each @xmath551 we have an isomorphism @xmath596 compose this isomorphism with the restrictions to respective boundary components and spectral projections to obtain the operator @xmath597 sending @xmath598 to @xmath599 here , we used the notation @xmath600 . note that @xmath601 is a smoothing operator on @xmath602 while @xmath603 is a smoothing operator on @xmath604
. the operator @xmath594 can now be written as @xmath605 sending @xmath606 to @xmath607 since the operators @xmath594 and @xmath608 are isomorphic , we again conclude that @xmath609 * step 5 . *
consider the last two components of @xmath608 , that is , the operator @xmath610 sending @xmath611 to @xmath612 this operator is obviously surjective .
therefore , we can apply lemma [ kernel to domain ] to the first four components of @xmath608 restricted to the kernel of the last two components .
the resulting operator @xmath613 sends @xmath614 to @xmath615 it follows from lemma [ kernel to domain ] that the operator @xmath608 is fredholm if and only if @xmath616 is fredholm , and @xmath617 * step 6 . *
the operator @xmath616 splits as @xmath618 , where the operator @xmath619 sends @xmath620 to @xmath621 and the operator @xmath62 sends @xmath622 to @xmath623 according to lemma [ assum ] , the operator @xmath600 has zero kernel .
denote by @xmath624 the smallest absolute value of the eigenvalues of @xmath542 then the operator norm of @xmath62 does not exceed @xmath625 , where @xmath626 is a constant independent of @xmath267 .
therefore , if @xmath267 is sufficiently large , the operator @xmath616 is fredholm if @xmath619 is fredholm , and in this case @xmath627 the operator @xmath619 further splits as a direct sum of the dirac operator with the atiyah patodi
singer boundary conditions , @xmath628 and an infinite family of operators @xmath629 for @xmath630 . by lemma [ assum ] , each of the operators @xmath631 with @xmath551 is an isomorphism .
therefore , the operator @xmath619 is fredholm if and only if @xmath632 is fredholm , and @xmath633 the operator @xmath632 is precisely the dirac operator @xmath634 with the atiyah patodi
singer boundary conditions .
since @xmath574 , the operator @xmath635 is a fredholm operator of index @xmath636 see atiyah
patodi singer ( * ? ? ?
* proposition 3.11 ) .
this completes the proof of theorem [ thm ] .
in this section , we continue the study of manifolds @xmath506 defined in by stretching the neck of a spin manifold @xmath2 of dimension @xmath500
. we will be interested in estimating the first eigenvalue of @xmath637 as @xmath28 .
this estimate will be used in the compactness argument in section [ s : compact ] .
[ eigenvalue estimate ] let us assume that the spin dirac operator @xmath638 is an isomorphism .
then there exist constants @xmath639 and @xmath640 such that for any @xmath641 , the operator @xmath642 has no eigenvalues in the interval @xmath643 . for the purpose of this proof
, we will view @xmath15 as a cobordism from @xmath37 to @xmath314 with @xmath644 .
the manifold @xmath645\times y)$ ] is then obtained from @xmath15 by gluing @xmath646 to @xmath314 and @xmath647 to @xmath401 .
denote by @xmath648 and @xmath649 the subspaces spanned by the eigenspinors of @xmath650 and @xmath651 with respectively positive and negative eigenvalues .
the @xmath29 orthogonal projections onto these subspaces will be denoted by @xmath544 , and the restriction maps @xmath652 will be denoted by @xmath653 with @xmath654 . as in lemma
[ assum ] , the operator @xmath513 is an isomorphism if and only if the following two conditions are satisfied : 1 .
the dirac operator @xmath542 has zero kernel , and 2 . the dirac operator @xmath655 with the atiyah patodi
singer boundary conditions is an isomorphism .
the operator @xmath656 is a non - negative self - adjoint elliptic differential operator hence all of its eigenvalues have the form @xmath657 with a real @xmath658 .
using the fact that @xmath659 has zero index , one can easily check that @xmath660 is an eigenvalue of @xmath661 if and only if the operator @xmath662 has non - zero kernel .
we denote the restriction of @xmath663 to @xmath15 , respectively , @xmath260\times y$ ] by @xmath664 , respectively , @xmath665 . supposing that @xmath663 belongs to the kernel of @xmath666 , then the following conditions are satisfied : a. @xmath667 and @xmath668 on @xmath15 ; b. @xmath669 and @xmath670 on @xmath260 \times y$ ] ; c. @xmath671 ; d. @xmath672 [ from one end to the other ] ( 1 ) there exists a linear operator @xmath673 such that , for any @xmath665 satisfying ( ii ) , @xmath674 \(2 ) there exists a linear operator @xmath675 such that , for any @xmath665 satisfying ( ii ) , @xmath676 \(3 ) for any @xmath677 , there exist constants @xmath678 and @xmath679 such that , for any @xmath680 and @xmath681 , @xmath682 we focus on the case of @xmath683 since the other case is similar . over @xmath260 \times y$ ] , use clifford multiplication with @xmath684 to identify the bundles @xmath685 and @xmath686 with each other and with the pull back of the bundle @xmath135 .
this identifies the operators @xmath687\times y)$ ] and @xmath688\times y)$ ] with the operators @xmath689 and @xmath690 , respectively .
choose a complete system of orthonormal eigenspinors @xmath691 for @xmath542 , with corresponding eigenvalues @xmath692 , @xmath551 .
let @xmath693 ( respectively , @xmath694 ) denote the vector space spanned by @xmath691 , treated as a section over @xmath401 ( respectively , @xmath695 ) .
it is sufficient to define @xmath683 on each of the spaces @xmath696 with @xmath697 , which we will do next .
let us write @xmath698 and @xmath699 then condition ( ii ) takes the form @xmath700 for all @xmath242 and @xmath701 $ ] .
the matrix of this system will be denoted by @xmath702 .
recall that @xmath695 is identified with @xmath703 and @xmath401 is identified with @xmath647 , and express @xmath704 in terms of @xmath705 .
the computation that follows is elementary if a bit tedious .
the eigenvalues of @xmath702 are @xmath706 , where @xmath707 , corresponding to the eigenvectors @xmath708 .
the solutions of ( [ ode ] ) are explicitly given by the formula @xmath709 from which we obtain @xmath710 and @xmath711 therefore , @xmath712 where @xmath713 we define @xmath714 on each of the spaces @xmath715 by the respective matrix @xmath716 .
to derive estimate ( 3 ) we let @xmath717 be the smallest absolute value of the eigenvalues of the operator @xmath542 .
note that @xmath718 is positive by our assumption on the kernel of @xmath542 and that @xmath719 for all @xmath242 . for any @xmath720 we obviously have @xmath721 hence @xmath722 and @xmath723 therefore , the norms of @xmath724 approach zero uniformly over @xmath242 as @xmath28 and @xmath725 .
this proves claim ( 3 ) for the operator @xmath683 .
we now return to the proof of proposition [ eigenvalue estimate ] .
it follows from lemma [ from one end to the other ] together with conditions ( iii ) and
( iv ) that @xmath726 therefore , the pair @xmath664 belongs to the kernel of the operator @xmath727 , where the operators @xmath728 are given by the formulas @xmath729 one can easily see that @xmath730 is isomorphic to the operator @xmath731 hence its kernel is zero by our assumption on the kernel of @xmath579 ( note that the operator @xmath579 has zero index ) .
therefore , there exists a constant @xmath732 such that the operator @xmath733 has zero kernel as long as @xmath734 , and so does the operator @xmath580 .
the proposition now follows from lemma [ from one end to the other ] ( 3 ) .
the following result , which will be used in section [ s : eta ] , is a straightforward extension of proposition [ eigenvalue estimate ] to the holomorphic family of operators @xmath735 where @xmath736 is an arbitrary smooth function such that @xmath737 = \gamma \in h^1 ( x;\mathbb z)$ ] .
note that the operators @xmath738 and @xmath739 are adjoint to each other whenever @xmath740 .
[ eigenvalue estimate_z ] let us assume that the spin dirac operator @xmath638 is an isomorphism .
then there exist constants @xmath678 and @xmath741 such that for any @xmath742 , the operators @xmath743 have no eigenvalues in the interval @xmath744 . the above proof can easily be adapted by introducing an extra parameter @xmath745 into the matching of the spinor bundles over @xmath37 .
this preserves the conditions ( i ) and ( ii ) but replaces the conditions ( iii ) and ( iv ) with * @xmath746 ; * @xmath747 .
the new operators @xmath714 and @xmath748 that show up in the formula for @xmath62 are obtained from the old ones by multiplying them on the left by @xmath749 respectively . since @xmath740 , this does not change the operator norm of @xmath62 , and the rest of the proof goes through with no change .
the proof of theorem [ t : sw = lefschetz ] naturally divides into two steps : compactness and gluing . in this section , we provide the necessary compactness results ; the proof of theorem [ t : sw = lefschetz ] will be completed in section [ s : gluing ] .
let @xmath2 be a connected smooth spin compact 4-manifold with a primitive cohomology class @xmath501 .
we will assume that the poincar dual of @xmath56 is realized by a rational homology 3-sphere @xmath14 and choose a metric @xmath91 on @xmath2 which takes the form @xmath254 in a product region @xmath255
\times y$ ] .
given a real number @xmath750 , consider the manifold with long neck @xmath751 \times y)\ ] ] obtained by cutting @xmath2 open along @xmath259 and gluing in the cylinder @xmath752 \times y$ ] along the two copies of @xmath9 .
this differs from the notation @xmath506 used in section [ s : correction ] by a simple re - parametrization .
in addition , for any @xmath753 we will write @xmath754 , where @xmath755\times y ) \cup w \cup ( [ 0,t ' ] \times y)\quad\text{and}\quad i_{t',t } = [ -t+t',t - t']\times y.\ ] ] we will find it convenient to extend these notations to the case of @xmath756 by letting @xmath757 be the disjoint union @xmath758 \times y \cup [ 0,+\infty ) \times y$ ] and using @xmath759 to denote the manifold @xmath26 with infinite product ends , as in section [ s : correction ] .
when @xmath756 and @xmath760 is finite , the notation @xmath761 will mean @xmath762 .
recall that , in order to define the monopole floer homology of a rational homology sphere @xmath9 in section [ s : hm ] , we introduced perturbations @xmath163 .
we will assume that our perturbation @xmath163 satisfies assumption [ small perturbation ] with respect to a constant @xmath173 satisfying @xmath763 , where @xmath640 is the constant from proposition [ eigenvalue estimate ] . to define the morphisms induced on the floer homology of @xmath9 by the spin cobordism @xmath15
, we will need to introduce further perturbations . to this end , consider a collar neighborhood @xmath764\times \partial w$ ] , with @xmath765 identified with the actual boundary @xmath766 .
let @xmath767 be a cut - off function which equals @xmath342 near @xmath768 and equals @xmath769 near @xmath770 , and let @xmath771 be a bump function with compact support in @xmath772 .
pick another perturbation @xmath773 as in section [ s : hm ] , and let @xmath774 where @xmath775 and @xmath776 are the 4-dimensional perturbations corresponding to @xmath163 and @xmath773 respectively ; see ( * ? ? ?
* definition 10.1.1 ) .
this is a perturbation on @xmath15 , supported in @xmath218 . by gluing the perturbations @xmath777 on @xmath15 and @xmath775 on @xmath778 together ,
we obtain a perturbation @xmath779 on @xmath780 .
similarly , we define a perturbation @xmath781 on @xmath26 by gluing together the perturbations @xmath777 on @xmath15 and @xmath775 on @xmath782 .
these perturbations give rise to the perturbed seiberg
witten equations , whose solutions will be referred to as monopoles .
let @xmath783 be the perturbed chern
dirac functional as in section [ s : hm ] .
downstairs , the gauge equivalence classes of its critical points form the finite set @xmath784\ ] ] where @xmath184 $ ] is the unique reducible class and @xmath785 consists of the irreducible classes . given @xmath196 , [ \beta ] \in \widetilde{\mathfrak{c}}$ ] , consider the following moduli spaces : 1 .
the moduli space @xmath786,[\beta])$ ] of unparameterized ( downstairs ) trajectories of the perturbed chern simons dirac gradient flow ( that is , monopoles on @xmath508 ) limiting to @xmath196 $ ] and @xmath200 $ ] at minus and plus infinity , in other words , the quotient of @xmath787,[\beta])$ ] by translations , excluding the constant trajectory if @xmath196 = [ \beta]$ ] , and 2 .
the moduli space @xmath788 , [ \beta])$ ] of ( downstairs ) monopoles on @xmath263 limiting to @xmath196 $ ] and @xmath200 $ ] at minus and plus infinity . we will write @xmath789)$ ] for @xmath789 , [ \alpha])$ ] .
since @xmath163 is admissible by assumption [ small perturbation ] , the moduli space @xmath786,[\beta])$ ] is always regular .
the regularity of the moduli space @xmath790 , [ \beta])$ ] is proved in the following lemma .
[ l : p_0 ] for any nice admissible perturbation @xmath163 , there exists a nice perturbation @xmath773 such that , for the perturbation ( [ e : mixed perturbation ] ) , the following conditions are satisfied : * the various moduli spaces of upstairs monopoles on @xmath263 are all regular . as a result , the cobordism induced map @xmath791 can be defined using this perturbation ; * the moduli space @xmath788 , [ \beta])$ ] is regular for all @xmath196,[\beta]\in \widetilde{\mathfrak{c}}$ ] .
furthermore , we may assume that @xmath773 is nice and that it satisfies the estimate @xmath792 for any @xmath177 , where @xmath793 is the product connection and @xmath171 is the constant fixed in the beginning of this section .
the proof is a careful check that the arguments of ( * ? ? ?
* proposition 24.4.7 ) hold in our situation .
we first introduce a large banach space @xmath794 of _ nice _ perturbations and form the parametrized moduli space @xmath795 after proving the regularity of @xmath796 , we can apply the sard smale lemma to find a residual subset @xmath797 with the property that @xmath798 is regular for any @xmath799 . in particular , we can choose a @xmath799 satisfying the estimate ( [ e : derivative of perturbation small 2 ] ) .
there is one new feature in this argument : since @xmath794 only consists of nice perturbations , at a reducible monopole , we can only obtain the transversality in the spinor direction by repeating the arguments in @xcite .
this does not cause a problem for the following reason : at a reducible monopole , the linearization of the curvature equation @xmath800 is the operator @xmath801 since @xmath802 , this operator is surjective ( without any perturbation ) . as a result ,
the transversality in directions tangent to the space of connections is automatically satisfied .
from now on , we will fix a perturbation @xmath163 satisfying assumption [ small perturbation ] and a perturbation @xmath803 as in lemma [ l : p_0 ] .
the following compactness theorem is the main result of this section . [ compactness ]
let @xmath804 be a sequence of positive real numbers such that @xmath805 .
then for any sequence @xmath806\in { \mathcal m}(x_{t_{n}})$ ] there exist @xmath196\in \mathfrak{c}^{*}$ ] and @xmath807\in { \mathcal m}(w_{\infty } , [ \alpha])$ ] such that , after passing to a subsequence , @xmath806 $ ] converges to @xmath807 $ ] in the sense of definition [ convergence ] below .
[ convergence ] let @xmath804 be a sequence of positive real numbers such that @xmath805 .
we will say that @xmath806\in { \mathcal m}(x_{t_{n}})$ ] converges to @xmath807\in { \mathcal m}(w_{\infty},[\alpha])$ ] if the following two conditions hold : 1 .
there exists a sequence of @xmath808 gauge transformation @xmath809 such that @xmath810 where by @xmath811 convergence we mean @xmath812 convergence on compact subsets of @xmath263 , and 2 . for any @xmath813 ,
there exist a real number @xmath750 and an integer @xmath814 such that , for all @xmath815 , we have @xmath816 and there is a sequence of @xmath808 gauge transformations @xmath817 such that @xmath818 here , @xmath819 stands for the constant trajectory of the chern simons
dirac gradient flow ( that is , a translation invariant monopole on the cylinder @xmath820 ) connecting the critical point @xmath18 to itself .
condition ( 2 ) roughly says that , up to a gauge transformation , @xmath821 is a `` near - constant trajectory '' when restricted to the middle of the long neck .
note that the definition of convergence given on page 486 of @xcite only includes condition ( 1 ) .
actually , condition ( 2 ) follows from condition ( 1 ) in our case but proving this would require some additional work . instead of doing this , we simply include condition ( 2 ) in our definition of convergence . the rest of this section will be dedicated to the proof of theorem [ compactness ] .
we begin with some preparations .
the perturbed topological energy of a configuration @xmath822 on a @xmath1-manifold was defined by kronheimer and mrowka in ( * ? ? ?
* definition 24.6.3 ) . on a cylinder
@xmath823\times y$ ] , the perturbed topological energy of @xmath822 is given by @xmath824 more generally , the topological energy of a configuration @xmath822 on the cobordism @xmath825 is given by @xmath826 [ energy bound ] ( 1 ) let @xmath822 be a monopole on a cylinder @xmath823\times y$ ] then @xmath827 , with equality if and only if @xmath822 is gauge equivalent to a constant trajectory .
\(2 ) there exists a constant @xmath626 such that @xmath828 for any monopole @xmath822 on @xmath825 .
claim ( 1 ) is clear because any monopole on a cylinder is gauge equivalent to a downward gradient flow line of @xmath829 .
claim ( 2 ) is a direct consequence of ( * ? ? ?
* lemma 24.5.1 ) and the equality of the topological and analytic energies for monopoles , see ( * ? ? ?
* definition 4.5.4 ) .
[ local compactness ] ( 1 ) let @xmath821 be a sequence of monopoles on @xmath823 \times y$ ] satisfying a uniform bound @xmath830 . then ,
after passing to a subsequence , there exist gauge transformations @xmath831\,\times\,y\rightarrow s^{1}$ ] and a monopole @xmath832 on @xmath823\,\times\,y$ ] such that the sequence @xmath833 converges to @xmath832 in the @xmath812 norm on every interior domain in @xmath823\times y$ ] .
\(2 ) let @xmath834 $ ] be a sequence of intervals with @xmath835 and @xmath836 , and @xmath821 a sequence of monopoles on @xmath837 satisfying a uniform bound @xmath830 . then ,
after passing to a subsequence , there exist gauge transformations @xmath838\,\times\,y\to s^{1}$ ] and a monopole @xmath807 \in { \mathcal m}([\alpha],[\beta])$ ] with @xmath196 , [ \beta]\in \mathfrak{c}$ ] such that the sequence @xmath833 converges to @xmath832 in @xmath839 .
\(3 ) let @xmath840 be a sequence of positive real numbers with @xmath841 , and @xmath821 a sequence of monopoles on @xmath842 satisfying a uniform bound @xmath830 . then , after passing to a subsequence , there exist gauge transformations @xmath843 and a monopole @xmath807\in { \mathcal m}({w_{\infty}},[\alpha],[\beta])$ ] with @xmath196 , [ \beta]\in \mathfrak{c}$ ] such that the sequence @xmath833 converges to @xmath832 in @xmath811 .
this follows from ( * ? ? ?
* theorem 10.7.1 and theorem 24.5.2 ) . for every @xmath196\in \widetilde{\mathfrak{c}}$ ]
, choose an open neighborhood @xmath844}\subset { \mathcal b}(y)$ ] such that @xmath844}\cap u_{[\beta ] } = \emptyset$ ] when @xmath196\neq [ \beta]$ ] .
in addition , for every @xmath196 \in \widetilde{\mathfrak c}$ ] , choose an open neighborhood @xmath845}$ ] of the constant trajectory @xmath846\in { \mathcal b}([0,1 ] \times y)$ ] such that @xmath847\in u_{[\alpha]}$ ] for all @xmath848 $ ] and all @xmath849\in u_{[\gamma_{\alpha}]}$ ] . here
\times y)$ ] denotes the space of gauge equivalent classes of @xmath812 configurations over @xmath851\times y$ ] .
[ small energy near constant ] there exists a constant @xmath171 such that for any monopole @xmath822 on @xmath852\times y$ ] of energy @xmath853 we have @xmath854\times y}]\subset u_{[\gamma_{\alpha}]}$ ] for some @xmath196\in \widetilde{\mathfrak c}$ ] .
suppose that this is not true .
then we can find a sequence of monopoles @xmath821 on @xmath852 \times y$ ] such that @xmath855 as @xmath856 but @xmath857\times y}]$ ] does not belong to any @xmath845}$ ] . by lemma [ local compactness ] ( 1 ) , after passing to a subsequence , @xmath857\times y}]$ ] will converge to a monopole of zero energy , which by lemma [ energy bound ] ( 1 ) must be of the form @xmath846 $ ] for some @xmath196\in \mathfrak{c}$ ] .
this leads to a contradiction .
[ the middle of a long neck ] for any real numbers @xmath858 and @xmath813 there exists a constant @xmath859 with the following significance : let @xmath860 and let @xmath822 be a monopole on @xmath752 \times y$ ] such that @xmath861 \in u_{[\alpha]}\quad\text{for all}\quad t\in [ -t , t].\ ] ] then for any interval @xmath862 \subset [ -t+t',t - t']$ ] , there exists a gauge transformation @xmath863\,\times\ , y\to s^1 $ ] such that @xmath864\times y}-\gamma_{\alpha}\ , \right\|_{l^2_k([t , t+4]\times y ) } < \epsilon.\ ] ] suppose that this is not true .
then there exist a sequence @xmath865 and a sequence @xmath821 of monopoles on @xmath866 \times y$ ] with @xmath867 such that @xmath868 and @xmath869 \in u_{[\alpha]}$ ] for all @xmath870 $ ] but , for some interval @xmath871 \subset [ -t_n + t'_n , t_n - t'_n]$ ] , we have @xmath872\times y}-\gamma_{\alpha}\,\right\|_{l^2_k([t_n , t_n+4]\times y)}\;\geq\;\epsilon\ ; > 0\ ] ] for all @xmath64 and all gauge transformations @xmath873\,\times\ , y\to s^1 $ ] . using the translation invariance of the seiberg witten equations on the cylinder , re - parameterize @xmath874 to obtain a monopole on @xmath875 $ ] , called again @xmath876 , such that @xmath877\times y}-\gamma_{\alpha}\,\right\|_{l^2_k([0,4]\times y)}\;\geq\;\epsilon\ ; > 0\ ] ] for all @xmath64 and all gauge transformations @xmath878\,\times\ , y\to s^1 $ ] .
note that @xmath879 and @xmath880 .
therefore , by lemma [ local compactness ] ( 2 ) , after passing to a subsequence , we can find gauge transformations @xmath881 \times y \to s^{1}$ ] such that @xmath882 in @xmath883 , where @xmath832 is a monopole on @xmath884 limiting to critical points in @xmath885 at plus and minus infinity .
since the gauge equivalence class of the restriction of @xmath832 to each slice @xmath886 is contained in @xmath844}$ ] , and since @xmath196 $ ] is the only critical point in @xmath844}$ ] , the monopole @xmath832 much be gauge equivalent to @xmath819 ( note that if @xmath9 were not a rational homology sphere , we would need to impose the extra condition that @xmath844}$ ] is contractible ) . without loss of generality
, we may assume that @xmath887 .
but then the @xmath883 convergence implies that @xmath888\times y}-\gamma_{\alpha}\,\right\|_{l^2_k([0,4]\times y)}\to 0,\ ] ] which leads to a contradiction .
[ global norm estimate ] for any @xmath813 there exists @xmath889 with the following significance : for any @xmath890 and any irreducible critical point @xmath196\in
\mathfrak c^*$ ] , let @xmath822 be a monopole on @xmath752\times y$ ] with the property that , for any @xmath862 \subset [ -t , t]$ ] , there exist a gauge transformation @xmath891\times y\rightarrow s^1 $ ] such that @xmath892\times y}-\gamma_{\alpha}\|_{l^2_k([t , t+4]\times y)}\leq \delta.\ ] ] then there exists a gauge transformation @xmath893 \times y\rightarrow s^1 $ ] such that @xmath894\times y}-\gamma_{\alpha}\|_{l^2_k([-t+5,t-5]\times y)}\leq \epsilon.\ ] ] this is essentially lemma 19.3.2 of @xcite .
in fact , the patching argument in the proof of lemma [ converging to reducible ] will be extracted , as in the proof of this lemma , from ( * ? ? ? * lemma 13.6 . ) .
let @xmath895 be the absolute grading function @xcite on the irreducible critical points , and extend it to the unique reducible @xmath184 $ ] by the formula @xmath896)\ , = -2 n(y , h,{\mathfrak s})\ ] ] ( compare with the grading in lemma [ l : reducible grading ] ) .
then , for any @xmath196 , [ \beta]\in \mathfrak c^*$ ] , the expected dimensions ( denoted by @xmath897 ) of the moduli spaces are as follows : 1 .
@xmath898,[\beta]))\ , = \ , { \operatorname{gr}}^{\mathbb{q}}([\alpha ] ) - { \operatorname{gr}}^{\mathbb{q } } ( [ \beta ] ) - 1 $ ] ; 2 .
@xmath898,[\,\theta]))\ , = \ , { \operatorname{gr}}^{\mathbb{q}}([\alpha ] ) - { \operatorname{gr}}^{\mathbb{q}}([\theta ] ) - 1 $ ] ; 3 .
@xmath899,[\alpha]))\ , = \ , { \operatorname{gr}}^{\mathbb{q}}([\theta ] ) - { \operatorname{gr}}^{\mathbb{q}}([\alpha ] ) - 2 $ ] ; 4 .
@xmath900,[\beta ] ) ) = { \operatorname{gr}}^{\mathbb{q}}([\alpha ] ) - { \operatorname{gr}}^{\mathbb{q}}([\beta])$ ] ; 5 . @xmath900,[\,\theta ] ) ) = { \operatorname{gr}}^{\mathbb{q}}([\alpha ] ) - { \operatorname{gr}}^{\mathbb{q}}([\theta])$ ] ; 6 .
@xmath901,[\alpha ] ) ) = { \operatorname{gr}}^{\mathbb{q}}([\,\theta ] ) - { \operatorname{gr}}^{\mathbb{q}}([\alpha])-1 $ ] ; 7 . @xmath901,[\theta]))=-1 $ ] . by our regularity assumption , the actual dimensions of the moduli spaces are equal to their expected dimensions except in case ( 7 ) : in this case , there is always one dimensional cokernel of the corresponding fredholm operator . as a result ,
@xmath902,[\theta])$ ] is a zero - dimensional manifold which only contains reducible monopoles .
note that the moduli space @xmath903,[\theta])$ ] is empty because we only allow non - constant trajectories in our definition of the moduli spaces @xmath904,[\alpha])$ ] .
[ broken at most once ] let @xmath905 and suppose that @xmath906 $ ] , , @xmath907 $ ] are critical points such that the moduli spaces @xmath908,[\alpha_2])$ ] ,
, @xmath909,[\alpha_{d-1}])$ ] and @xmath910,[\alpha_{1}])$ ] are all non - empty . then @xmath911 .
this follows from our regularity assumption by a simple dimension count .
we will follow closely the argument of ( * ? ? ?
* section 16.2 ) .
let @xmath876 be a sequence of monopoles as in the statement of theorem [ compactness ] .
since @xmath912 \times y}\right ) = 2\left({\mathcal l}_{\mathfrak{q}}\left((a_n,{\varphi}_n)|_{\{-t_n\}\times y}\right ) - { \mathcal l}_{\mathfrak{q}}\left((a_n,{\varphi}_n)|_{\{t_n\}\times
y}\right)\right ) \\ = - { \mathcal e}_{\mathfrak{q}}^{{\operatorname{top}}}\left((a_n,{\varphi}_n)|_w\right),\end{gathered}\ ] ] we conclude from lemma [ energy bound ] that @xmath913 \times y}\right)\,\leq m\ ] ] for some constant @xmath914 independent of @xmath64 . as a result , we obtain the uniform bounds @xmath915 for any @xmath916 and any interval @xmath917 $ ] , which will allow us to apply lemma [ local compactness ] .
choose neighborhoods @xmath844}$ ] and @xmath845}$ ] as in section [ near constant trajectories ] , and let @xmath171 be the constant provided by lemma [ small energy near constant ] .
restricting @xmath876 to the slices @xmath886 gives rise to a path @xmath918 \to { \mathcal b}(y)$ ] .
for each @xmath64 , consider the set @xmath919 \subset [ -t_n , t_n]\;\;\text{and}\;\;{\mathcal e}_{\mathfrak{q}}^{{\operatorname{top}}}\left((a_n,{\varphi}_n)|_{[p , p+1]\times y}\right)\geq \epsilon_0 > 0\,\right\}.\ ] ] this set contains at most @xmath920 elements . by passing to a subsequence
, we may find an integer @xmath451 such that every set @xmath921 contains exactly @xmath451 elements , @xmath922 . also introduce the integers @xmath923 and @xmath924 .
after passing to a subsequence one more time , we may assume that , for each integer @xmath483 between @xmath769 and @xmath451 , either @xmath925 or @xmath926 is independent of @xmath64 . on the set @xmath927 , define an equivalence relation @xmath928 denote by @xmath929 the number of the equivalence classes ; since @xmath769 is not equivalent to @xmath451 , we must have @xmath905 .
pick representatives @xmath930 , one for each equivalence class , and let @xmath931 then @xmath932 the difference @xmath933 is independent of @xmath64 for @xmath934 , and @xmath935 for @xmath936 .
using lemma [ small energy near constant ] together with the translation invariance of the seiberg
witten equations , we obtain @xmath937)\subset u_{[\alpha_{m+1}]}\ ] ] for some critical point @xmath938 $ ] . here
, we passed to a subsequence again to ensure that @xmath938 $ ] is independent of @xmath64 . using lemma [ local compactness ] and passing to a subsequence if necessary we conclude the following : 1
there exist gauge transformations @xmath939 such that @xmath940 converges in @xmath941 to a monopole @xmath832 on @xmath263 . using ( [ eq 1 ] ) , it is not difficult to see that @xmath807 \in { \mathcal m}({w_{\infty}},[\alpha_{d-1}],[\alpha_{1}])$ ] .
2 . let @xmath942 denote the translation on trajectories defined by @xmath943 .
then , for every @xmath944 and @xmath945 , we have @xmath946 where @xmath947 is a trajectory on @xmath948 .
since the topological energy of @xmath949)$ ] is greater than or equal to @xmath171 , we conclude that @xmath947 is not a constant trajectory . using ( [ eq 1 ] )
, it is not difficult to see that @xmath947 represents a monopole in @xmath950,[\alpha_{m+1}])$ ] . since all the moduli spaces @xmath951,[\alpha_{2 } ] ) , \cdots , { \mathcal m}([\alpha_{d-2}],[\alpha_{d-1}])$ ] and @xmath952,[\alpha_{1}])$ ] are non - empty , it follows from lemma [ broken at most once ] that @xmath911 . keeping this in mind ,
our earlier discussion implies the following two results : 1 .
there exist gauge transformations @xmath953 such that @xmath954 in @xmath941 for some @xmath807 \in { \mathcal m}({w_{\infty}},[\alpha])$ ] , where @xmath196 = [ \alpha_1]$ ] ; 2 .
there exists a constant @xmath859 such that @xmath955 ) \subset u_{[\alpha]}$ ] for all @xmath64 ( this is implied by ( [ eq 1 ] ) ) . using lemma [ the middle of a long neck ] , we can replace ( b ) by the following : 1 . for any @xmath813
, there exists a constant @xmath515 with the following significance : for any @xmath64 large enough so that @xmath956 and any interval @xmath957 \subset [ -t_n + c , t_n - c]$ ] , there exists a gauge transformation @xmath958\times y \to s^1 $ ] such that @xmath959\times y}-\gamma_{\alpha}\|_{l^2_k([t , t+4]\times y)}\,\leq\ , \epsilon.\ ] ] [ converging to reducible ] the critical point @xmath196 $ ] is irreducible .
this lemma will be proved in section [ s : reducible ] . for now , let us assume it and finish the proof of theorem [ compactness ] .
it is clear that condition ( 1 ) of definition [ convergence ] follows from ( a ) and lemma [ converging to reducible ] .
therefore , we only need to verify condition ( 2 ) of definition [ convergence ] . for any @xmath813 ,
let @xmath626 be the constant from ( c ) .
choose @xmath960 and let @xmath287 be an integer large enough so that @xmath816 for all @xmath961 .
a straightforward application of lemma [ global norm estimate ] finishes the proof . in this subsection
we prove lemma [ converging to reducible ] .
the proof will be based on the following three lemmas .
the moduli space @xmath902)$ ] contains a single point @xmath962 $ ] , where @xmath963 is a trivial connection on @xmath26 .
since both @xmath163 and @xmath773 are nice , we may disregard the perturbations when studying reducible monopoles downstairs .
we saw in section [ s : broken ] that the moduli space @xmath902)$ ] contains only reducible monopoles @xmath964 $ ] with @xmath965 .
write @xmath966 , where @xmath467 is an @xmath967 differential 1-form on @xmath26 with coefficients in @xmath968 satisfying @xmath969 .
integration by parts shows that @xmath970 therefore , the 1-form @xmath467 is closed .
since @xmath971 , there exists @xmath972 such that @xmath973 , and @xmath974 is gauge equivalent to @xmath975 via the gauge transformation @xmath976 .
now we use ( * ? ? ?
* definition 24.2.1 ) to conclude that @xmath964 = [ ( a_0,0)]$ ] .
[ l : four ] for any @xmath813 there exists a positive integer @xmath287 such that , for any @xmath977 , 1 .
@xmath978 2 .
there exists a gauge transformation @xmath979 such that @xmath980 3 . for any interval @xmath862 \subset
[ -t_n , t_n]$ ] there exists a gauge transformation @xmath958\times y \to s^1 $ ] such that @xmath981\times y}-(a_0,0)\|_{l^2_k([t , t+4]\times y ) } < \epsilon.\ ] ] claim ( 1 ) follows trivially from @xmath805 . since @xmath982 is a compact subset of @xmath26 , claim ( 2 ) follows from ( a ) . to prove claim ( 3 ) ,
let @xmath515 be the constant from ( c ) .
if @xmath862 $ ] belongs to @xmath983 $ ] , claim ( 3 ) follows from ( c ) .
otherwise , @xmath862 $ ] belongs to either @xmath984 $ ] or @xmath985 $ ] . for every @xmath64 ,
these are identified with the fixed compact subsets @xmath986 $ ] and @xmath987 $ ] of @xmath26 hence the result follows from ( a ) .
[ l : l2 norm of perturbation ] let @xmath171 be the constant fixed in the beginning of section [ s : perturbation ] . then there exists an integer @xmath814 such that , for any @xmath988 , we have @xmath989 since @xmath876 solves the perturbed seiberg
witten equations , we have the equality @xmath990 where @xmath991 denotes the spinor component of the perturbation term @xmath992 ; see section [ s : perturbation ] .
it is supported in @xmath993 , where @xmath218 is a collar neighborhood of @xmath994 . by our definition of @xmath995 , @xmath996.\
] ] since @xmath163 satisfies assumption [ small perturbation ] , it follows from ( [ e : derivative of perturbation small ] ) that there exists a neighborhood @xmath472 of @xmath184\in { \mathcal b}(y)$ ] such that @xmath997 for any configuration @xmath140 with @xmath998\in v$ ] .
but then , by lemma [ l : four ] , there exists a positive integer @xmath287 such that for any @xmath988 , we have @xmath999\in v$ ] for all @xmath1000 $ ] .
therefore , we have the estimate @xmath1001$},\ ] ] which implies that @xmath1002 a similar argument involving estimate shows that @xmath1003 this completes the proof of the lemma because @xmath1004 is supported on @xmath993 .
[ l : global ] for any @xmath813 there exists a positive integer @xmath287 such that , for any @xmath977 , there is a global gauge transformation @xmath1005 such that @xmath1006 before we go on to prove this lemma , we will show how it implies lemma [ converging to reducible ] .
we know from proposition [ eigenvalue estimate ] that as @xmath1007 the smallest eigenvalue of the operator @xmath1008 on @xmath1009 is bounded from below by @xmath1010 .
therefore , for all sufficiently large @xmath64 and all positive spinors @xmath142 , @xmath1011 on the other hand , consider the sequence of gauge transformations @xmath1012 from lemma [ l : global ] and the sequence of spinors @xmath1013 with @xmath1014 .
lemma [ l : l2 norm of perturbation ] implies that @xmath1015 we then conclude using lemma [ l : global ] that @xmath1016 for all sufficiently large @xmath64 , which gives a contradiction .
let @xmath287 be as in lemma [ l : four ] then , for any @xmath977 , there exists an integer @xmath1017 such that @xmath1018 . for @xmath1019 ,
we denote the gauge transformation @xmath1020 of lemma [ l : four ] by @xmath1021 .
we wish to glue the gauge transformations @xmath1022 together with the help of cutoff functions .
first , we pick a base point @xmath1023 . after multiplying by suitable constant gauge transformations ,
we may assume that @xmath1024 since @xmath9 is a rational homology sphere , we have @xmath1025\times y}=e^{i\xi(n , j)}\cdot u(n , j+1)|_{[-t_{n}+2j+2,-t_{n}+2j+4]\times y},\ ] ] where @xmath1026\times y\rightarrow \mathbb{r}\ ] ] satisfies @xmath1027 now , since @xmath1028\times y ) } < \epsilon\ ] ] and @xmath1029\times y ) } < \epsilon,\ ] ] we have @xmath1030\times y ) } = \\ \|u(n , j)^{-1}du(n , j)-u(n , j+1)^{-1}du(n , j+1)\|_{l^{2}_{k}([-t_{n}+2j+2,-t_{n}+2j+4]\times y ) } < 2\epsilon.\end{gathered}\ ] ] together with ( [ base point normalization ] ) , this implies that there exists a constant @xmath1031 such that @xmath1032\times y)}\ ; < \ ; c_2\cdot \epsilon.\ ] ] next , choose a bump function @xmath1033\rightarrow [ 0,1]$ ] such that @xmath1034)=1 $ ] and @xmath1035)=0 $ ] .
we let @xmath1036\times y\rightarrow [ 0,1]$ ] be the function defined by the formula @xmath1037 . extend the function @xmath1038 defined on @xmath1039\times y\ ] ] by zero to obtain a function @xmath1040\times y\rightarrow \mathbb{r}.\ ] ] we have @xmath1041 ) } < c_3\ , \epsilon$ ] for a constant @xmath1042 , which implies that @xmath1043 \times y)}\ ; < \ ; c_4\cdot \epsilon\ ] ] for another constant @xmath1044 .
now , for @xmath1045 , consider the gauge transformations @xmath1046\times y\rightarrow s^{1}.\ ] ] one can easily see that @xmath1047 equals @xmath1048 over @xmath1049\times y$ ] and equals @xmath1021 over @xmath1050\times y$ ] .
therefore , the functions @xmath1051 , @xmath1052 , together with the functions @xmath1053\times y}$ ] and @xmath1054\times y}$ ] , agree with each other on the overlaps . as a result
, we can glue them together to obtain a gauge transformation @xmath1055\times y\rightarrow s^{1}.\ ] ] since @xmath1056 is obtained from @xmath1057 by multiplying by @xmath1058 , there is a constant @xmath1059 such that @xmath1060\times y)}\ ; < \ ; c_5\cdot \epsilon.\ ] ] in our next step , we introduce @xmath1061 , where @xmath1062 is the coordinate in the cylindrical direction , and @xmath1063 $ ] and @xmath1064 $ ] are chosen so that @xmath1065 and @xmath1066 there is a constant @xmath1067 such that @xmath1068\times y)}\;\leq\ ; c_{6}\cdot \epsilon + \pi/(t_{n}-2),\ ] ] where , by choosing @xmath814 large enough , we may assume @xmath1069 for all @xmath988 . arguing as before , we can find @xmath1070\cup [ -t_n+2m_n+2 , -t_n+2m_n+4])\times y\rightarrow \mathbb r\ ] ] such that @xmath1071 equals @xmath1072 over the domain of @xmath1073 and @xmath1074 as before , we have the estimate @xmath1075 on the domain of @xmath1076 .
let @xmath1077 be a cut - off function on the domain of @xmath1076 which equals @xmath342 when restricted to @xmath1078\cup [ -t_{n}+2m_{n}+2 , -t_{n}+2m_{n}+3])\times y$ ] and equals @xmath769 when restricted to @xmath1079\cup [ -t_{n}+2m_{n}+7/2 , -t_{n}+2m_{n}+4])\times y$ ] .
assume that the @xmath1080 norm of @xmath1077 is uniformly bounded for all @xmath988 and extend @xmath1081 by zero to a function @xmath1082 .
the gauge transformations @xmath1083 and @xmath1084\times y}$ ] match on the overlap of their domains , therefore , we can glue them together to the desired gauge transformation @xmath1085 .
since @xmath1086 is obtained by modifying @xmath1072 and @xmath1087 using the cutoff functions @xmath1088 , @xmath1089 and the function @xmath1090 the estimate of the lemma can be easily verified .
in this section , we will finish the proof of theorem [ t : sw = lefschetz ] by first establishing a bijective correspondence between monopoles on @xmath506 and monopoles on @xmath26 for all sufficiently large @xmath267 , and then matching the signs to identify @xmath1091 with the lefschetz number in the monopole chain complex @xmath1092 . to simplify notations
, we will continue writing @xmath1093 for @xmath1094 .
[ sw = lefschetz ] assume that the spin dirac operator @xmath1095 is an isomorphism .
then , for all sufficiently large @xmath257 , the moduli space @xmath1093 is regular , and there exists a homeomorphism @xmath1096\in\,\mathfrak{c}^ * } { \mathcal m}(w_{\infty } , [ \alpha]).\ ] ] we will first prove theorem [ sw = lefschetz ] by adopting the gluing techniques from @xcite to the non - separating case at hand . theorem [ t :
sw = lefschetz ] will be proved at the end of this section .
we will be using notations from section [ s : notations ] .
denote by @xmath1097 the moduli space of irreducible monopoles on @xmath1098 .
it follows from proposition [ eigenvalue estimate_z ] that @xmath1099 for all sufficiently large @xmath750 .
the similarly defined moduli spaces @xmath1100 and @xmath1101 will be infinite dimensional hilbert manifolds because both @xmath1102 and @xmath1103 have non - empty boundary but we are not imposing any boundary conditions . by the unique continuation theorem ( * ? ? ?
* section 10.8 ) , the restriction of an irreducible monopole to the boundary is irreducible , therefore , for all @xmath753 we have well defined restriction maps @xmath1104 where @xmath159 consists of irreducible configurations in @xmath157 .
one can show that these maps are embeddings of hilbert manifolds
. it will be convenient to extend these notations to the case of @xmath756 . recall that in section [ s : notations ] we defined @xmath1105 .
let @xmath1106 be the moduli space of monopoles @xmath822 on @xmath761 satisfying @xmath1107\ ; = \ ; \lim\limits_{t \to -\infty}\left[(a,{\varphi})|_{\{t\}\times y}\right]\ ; \in\ ; \mathfrak c^*.\ ] ] then the restriction map @xmath1108 is still well defined and is an embedding of hilbert manifolds . to unify the notations ,
we will write @xmath1109 and @xmath1110\in \mathfrak{c}^ { * } } { \mathcal m}^*({w_{\infty } } , [ \alpha]).\ ] ] then , for all @xmath1111 , we have the following commutative diagram whose unmarked arrows are given by restriction to submanifolds @xmath1112(b)--(a ) node [ midway , above](textnode){$r^-_{t'}$\ ; } ; \draw[->](c)--(a ) node [ midway , below](textnode){$r^+_{t',t}$ } ; \draw[->](d)--(b ) node [ midway , above](textnode ) { } ; \draw[->](d)--(c ) node [ midway , above](textnode ) { } ; \end{tikzpicture}\ ] ] [ solution as fiber product ] for all @xmath1111 , the above diagram is a cartesian square , that is , @xmath1097 is homeomorphic to the fiber product @xmath1113 the proof is identical to that of ( * ? ? ? * lemma 19.1.1 ) , which deals with the separating case , and will be omitted .
note that some new issues would appear if we were to glue reducible monopoles in the blown - up moduli space but we do not deal with them here . in less formal terms ,
lemma [ solution as fiber product ] asserts that the moduli space @xmath1097 is the intersection of the moduli spaces @xmath1100 and @xmath1101 viewed as submanifolds of @xmath1114
. we will prove theorem [ sw = lefschetz ] by showing that all of these intersections occur in small neighborhoods of constant trajectories @xmath819 , where the intersection points for sufficiently large @xmath1115 can be matched with those for @xmath756 using implicit function theorem . the central role in our proof
will be played by the following theorem , which is a special case of ( * ? ? ?
* theorem 18.2.1 ) .
when @xmath1115 is finite , @xmath1116 will stand for @xmath752\times y$ ] and @xmath1117 for the corresponding restriction map .
[ near constant solution on finite cylinder ] there exists a constant @xmath1118 such that for all @xmath1119 and @xmath196 \in \mathfrak c^*$ ] , there exist smooth maps @xmath1120}(t,- ) : b([\alpha ] ) \to { \mathcal m}^ * ( i_t ) \\
u_{[\alpha]}(\infty,- ) : b([\alpha])\to { \mathcal m}^ * ( i_{\infty})\end{gathered}\ ] ] which are diffeomorphisms from an open neighborhood @xmath1121)\subset { \mathcal b}^*(y)$ ] , which is independent of @xmath1115 , onto neighborhoods of the constant solutions @xmath846 $ ] .
moreover , the maps @xmath1122,t}\ ; = \ ; r^+_t\circ u_{[\alpha ] } ( t,-)\ , : b([\alpha ] ) \to { \mathcal b}^*(y)\times { \mathcal b}^*(y)\ ] ] are smooth embeddings for all @xmath1123 $ ] , and we have a @xmath1124 convergence @xmath1122,t}\ ; \longrightarrow\;\mu_{[\alpha],\infty}\quad\text{as}\quad t \to \infty.\ ] ] finally , there exists a constant @xmath1125 , independent of @xmath1115 , such that the image of the map @xmath1126}(t,-)$ ] contains all the trajectories @xmath1127\in { \mathcal m}^ * ( i_t)$ ] with @xmath1128 . [ separating ] in addition , we will assume that , for all @xmath1129 $ ] and all @xmath196\neq [ \beta ] \in \mathfrak c^*$ ] , @xmath1130}(s,-))\cap \operatorname{im}(u_{[\beta]}(t,-))=\emptyset,\quad\operatorname{im}(\mu_{[\alpha]}(s,-))\cap \operatorname{im}(\mu_{[\beta]}(t,-))=\emptyset.\ ] ] [ long neck near constant ] let @xmath1125 be as in theorem [ near constant solution on finite cylinder ]
. then one can find constants @xmath1131 with the following significance : for any @xmath1132 $ ] , any element of @xmath1097 can be represented by a monopole @xmath822 such that @xmath1133 for some @xmath196\in \mathfrak c^*$ ] .
the case @xmath756 should be clear once we remember that , for @xmath756 , the notation @xmath1134 means @xmath1135 .
let us now assume that @xmath1136 and suppose to the contrary that the constants @xmath1137 and @xmath1138 do not exist .
then we can find two sequences of real numbers @xmath1139 , both going to infinity as @xmath856 , and a sequence of monopoles @xmath876 on @xmath1009 such that , for any gauge transformation @xmath1140 and any @xmath1141 , we have @xmath1142 for all @xmath64 . according to theorem [ compactness ] , after passing to a subsequence , we may assume that @xmath806 $ ] converges to an element in @xmath1143)$ ] for some @xmath1144\in\mathfrak c^*$ ] .
this leads to a contradiction with part ( 2 ) of definition [ convergence ] of convergence .
[ solution as fiber product ii ] let @xmath1137 and @xmath1138 be the constants from lemma [ long neck near constant ] .
then for any @xmath1145 $ ] , there exists a homeomorphism @xmath1146\in \mathfrak c^*}{\operatorname{fib}}\ , ( r^-_{t_2},\,\mu_{[\alpha],t - t_2})\ ] ] between @xmath1147 and a disjoint union of the fiber products .
moreover , the moduli space @xmath1147 is regular if and only if , for any @xmath196\in \mathfrak c^*$ ] , the images of the maps @xmath1148 and @xmath1149,t - t_2}$ ] intersect transversely in @xmath1150 . the first assertion is a straightforward corollary of theorem [ near constant solution on finite cylinder ] , lemma [ solution as fiber product ] , and lemma [ long neck near constant ] .
the second assertion is essentially ( * ? ? ?
* theorem 19.1.4 ) .
[ convergence on the boundary ] let @xmath1121 ) \subset { \mathcal b}^*(y)$ ] be an open neighborhood as in theorem [ near constant solution on finite cylinder ] , and let @xmath1151)$ ] be a sequence such that @xmath1149,t_n}(x_n)\to \mu_{[\alpha],\infty}(x)$ ] for some @xmath805 and some @xmath1152)$ ] .
then @xmath1153 in the topology of @xmath159 .
this will be clear once we recall the construction of the map @xmath1126}(t,-)$ ] from section 18.4 of @xcite .
the tangent space @xmath1154}{\mathcal b}^*(y)$ ] , denoted by @xmath1155 , has a decomposition @xmath1156 given by the spectral decomposition of the hessian of @xmath1157 .
we will denote by @xmath1158 the corresponding projections .
we will also identify small open balls @xmath1159 ) \subset { \mathcal k}$ ] of radius @xmath813 with open neighborhoods @xmath1121)\subset{\mathcal b}^*(y)$ ] . according to ( * ? ? ?
* section 18.4 ) , for any sufficiently small @xmath813 , there exist constants @xmath1160 and @xmath1118 with the following significance : for any @xmath1161 $ ] and any @xmath1162)\ , \subset\ , { \mathcal k}_{\mathfrak{q}}^+\oplus { \mathcal k}_{\mathfrak{q}}^-$ ] , there exists a unique @xmath1163\;\in\ ; ( r^+_t)^{-1}(b([\alpha])\times b([\alpha]))\;\subset\ ; { \mathcal m}^*(i_t)\ ] ] such that @xmath1164 and @xmath1165 ) ) = ( a , b)$ ]
. the map @xmath1126}(t,-)$ ] is then defined by the formula @xmath1166}(t,(a , b ) ) = [ \gamma^t_{(a , b)}].\ ] ] with this definition in place , write @xmath1167 and similarly @xmath1168
. then @xmath1149,t_n}(x_n ) = ( a_n,*,*,b_n)$ ] and @xmath1149,\infty}(x_{\infty } ) = ( a_{\infty},*,*,b_{\infty})$ ] , where we embedded @xmath1121)\times b([\alpha])$ ] into @xmath1169 .
from this we clearly see that @xmath1149,t_n}(x_n)\to \mu_{[\alpha],\infty}(x)$ ] implies @xmath1170 .
let @xmath1137 and @xmath1138 be the constants from lemma [ long neck near constant ] . using lemma
[ solution as fiber product ii ] and our regularity assumption on @xmath1171 we can claim that , for any @xmath196 \in \mathfrak c^*$ ] , the images of the maps @xmath1172 and @xmath1149,\infty}$ ] intersect each other transversely and we have a homeomorphism ( bijection ) @xmath1173\in
\mathfrak c^*}{\operatorname{fib}}\ , ( r^-_{t_2},\,\mu_{[\alpha],\infty}).\ ] ] by theorem [ near constant solution on finite cylinder ] , the maps @xmath1149,t}$ ] converge to @xmath1149,\infty}$ ] in the @xmath1124 topology as @xmath1174 .
the implicit function theorem now implies that there exists a constant @xmath1175 such that any @xmath1176,\infty})$ ] has an open neighborhood @xmath1177}$ ] with the following significance : for any @xmath1178 , the images of embeddings @xmath1172 and @xmath1149,t - t_2}$ ] intersect each other in exactly one point in @xmath1179 , and this intersection is transverse .
therefore , our proof will be finished once we prove the following lemma .
there exists a constant @xmath1180 such that , for any @xmath1181 and any element of @xmath1097 represented by @xmath1182\in \mathfrak c^ * } { \operatorname{fib}}\ , ( r^-_{t_2},\,\mu_{[\alpha],t - t_2}),\ ] ] there exists a point @xmath1176,\infty})$ ] such that @xmath1183 .
suppose to the contrary that this is not the case .
then there is a sequence @xmath1184,t_n}),\quad t_n \to \infty,\ ] ] representing monopoles @xmath876 on @xmath1009 such that @xmath1185 for any @xmath1186 in the fiber product @xmath1187,\infty})$ ] . by theorem [ compactness ] , after passing to a subsequence , we may assume that @xmath806 $ ] converges to a monopole @xmath1188 $ ] on @xmath759 .
represent the latter by @xmath1189 , \infty})$ ]
. then * @xmath1190 .
this follows from part ( 1 ) of definition [ convergence ] of convergence because @xmath1191 is a compact subset of @xmath263 and @xmath1192\quad\text{and}\quad y_{\infty } = [ ( a_{\infty},{\varphi}_{\infty})|_{w_{t_2}}].\ ] ] * @xmath1193 .
this follows from remark [ separating ] , which implies that @xmath1194 = [ \alpha]$ ] for all large enough @xmath64 , and from lemma [ convergence on the boundary ] applied to the convergent sequence @xmath1195,t_n}(x_n)\ , = \
, r^-_{t_2}(y_{n})\;\longrightarrow\ ; r^-_{t_2}(y_{\infty})\ , = \ , \mu_{[\alpha],\infty}(x_{\infty}).\ ] ] therefore , @xmath1196 for all sufficiently large @xmath64 .
this gives a contradiction .
all we need to do is compare the signs with which the monopoles corresponding to each other under the map are counted in @xmath1197 and @xmath1198 .
this is done in ( * ? ? ?
* proposition 3 ) under a different @xmath74 grading convention .
since our setting here is slightly different , we give an alternative argument using excision principle . in the product case , @xmath19 , the orientation transport argument of @xcite
( see also @xcite in the instanton setting ) can be used to show that @xmath1199 up to an overall sign which is independent of the choice of @xmath9 .
the sign is determined by the sign fixing condition @xmath1200 of ( * ? ? ?
* section 11.2 ) , where @xmath20 is the casson invariant normalized so that @xmath121 . to calculate the sign in , we will let @xmath19 , where @xmath1201 is the brieskorn homology sphere endowed with a natural metric @xmath48 realizing its thurston geometry , and compute @xmath1202 with respect to the product metric @xmath254 .
the correction term in this formula equals @xmath1203 where @xmath110 can be any smooth compact spin manifold with boundary @xmath9 .
let us choose @xmath110 to be the plumbed manifold with the intersection form isomorphic to @xmath1204 , where @xmath1205 is positive definite . according to ( * ? ? ?
* section 6 ) , the index of @xmath1206 vanishes , therefore , @xmath1207 . since @xmath1208
we conclude from formulas and that @xmath1209 .
this needs to be compared to the euler characteristic of the chain complex @xmath1092 .
the latter complex is known @xcite to have exactly two generators of the same grading .
note that @xmath1210 ) = -2w(x , g ) = -2.\ ] ] by ( * ? ? ?
* lemma 2.9 ) , the boundary map @xmath1211 must be non - zero .
therefore , both generators of @xmath191 must be of odd grading , and the overall sign in formula is a minus .
the general case now follows by using the excision principle for determinant bundles as in ( * ? ? ?
* section 25 ) .
this section contains two results about metrics with no harmonic spinors on two types of spin manifolds : compact manifolds with product regions and non - compact manifolds with cylindrical ends .
these results are similar to those of amman , dahl , and humbert @xcite and are proved by a modification of their argument .
let @xmath2 be a connected smooth spin compact manifold of dimension @xmath500 and @xmath736 a smooth map with the primitive cohomology class @xmath737 \in h^1 ( x ; { \mathbb z})$ ] .
let @xmath14 be a connected manifold poincar dual to @xmath737 $ ] , and introduce riemannian metrics @xmath48 on @xmath9 and @xmath91 on @xmath2 so that @xmath254 in a product region @xmath255 \times y \subset x$ ] . according to ( * ? ? ? * theorem 1.1 ) , the spin dirac operator @xmath137 is invertible in the sobolev @xmath29 completion for a generic choice of @xmath48 .
[ t : one ] let @xmath2 be a manifold as above with a product region @xmath255 \times y$ ] and suppose that the metric @xmath48 is such that @xmath1212 .
if @xmath2 is spin cobordant to zero then there exists a metric @xmath91 on @xmath2 such that @xmath254 in the product region and @xmath1213 is invertible in the sobolev @xmath29 completion .
we will use the method of transporting invertibility of the operator @xmath1213 across a spin cobordism as in @xcite but we will be more specific in choosing the cobordism . start with the manifold @xmath1214 with the product metric @xmath1215 and the spin dirac operator @xmath1216 .
since the operator @xmath137 is invertible , so is the operator @xmath1217 by a direct calculation .
we claim that there is a spin cobordism from @xmath1218 to @xmath2 that contains a product region
@xmath1219 \times y$ ] , and furthermore has handles of index at most @xmath1220 . to prove this ,
write @xmath1221 \times y)\quad\text{and}\quad x ' = v \cup_{\{-{\varepsilon},{\varepsilon}\}\times y } [ -{\varepsilon},{\varepsilon } ] \times y;\ ] ] see figure [ f : spin ] .
we assume of course that @xmath472 has a product structure as well .
let us consider @xmath1222 \times \{-{\varepsilon},{\varepsilon}\}\times y)\ , \cup\ , v\ ] ] where @xmath1223 is identified with @xmath1224 and @xmath1225 is identified with @xmath994 . after rounding corners
, @xmath218 is a spin manifold diffeomorphic to @xmath2 hence @xmath218 is the boundary of a connected spin manifold @xmath267 of dimension @xmath1226 .
view @xmath267 as a cobordism from @xmath472 to @xmath15 , relative to the product region @xmath851 \times \partial v$ ] , and give it a handle decomposition . at 150 -10 at 150 50 at 150 107 at 20 130 at 15 11 at 250 70 by surgeries on generators of @xmath1227 , preserving the spin condition
, we can make @xmath267 simply connected , and cancelling the @xmath769-handles and @xmath1228-handles , we may assume that all the handles have index between @xmath342 and @xmath64 . then a standard handle trading argument can be used to eliminate the handles of index @xmath342 and @xmath64 . gluing this to the product region @xmath1219
\times y$ ] gives the desired cobordism .
according to ( * ? ? ?
* theorem 1.1 ) , the manifold @xmath2 admits a metric @xmath91 such that @xmath1229 .
this explicitly constructed metric matches the original product metric on @xmath1218 away from arbitrarily thin tubular neighborhoods of the surgery spheres . by construction , these spheres lie outside of the product region , so that the metric remains a product there . since @xmath2 is spin cobordant to zero , we conclude that @xmath1230
. therefore , the operator @xmath1213 has zero index and must be invertible .
the existence of just a single metric as in theorem [ t : one ] implies that the set of such metrics is in fact generic ( that is , @xmath1231 open and @xmath1080 dense ) in the space of all metrics on @xmath2 with a fixed product metric on @xmath255 \times y$ ] .
the proof of ( * ? ? ?
* proposition 3.1 ) goes through after one notes that fixing a product metric on @xmath255 \times y$ ] defines a convex subset in the space of all metrics .
let @xmath2 be a manifold from the previous section with a product region @xmath255
\times y$ ] and a metric @xmath91 which in the product region takes the form @xmath254 .
cut @xmath2 open along @xmath1232 into a cobordism @xmath15 , and attach infinite product ends to @xmath15 .
this results in the non - compact manifold @xmath1233\times y)\,\cup\,w\ , \cup\ , ( [ 0,\infty ) \times y).\ ] ] [ t : two ] let us suppose that @xmath1212 and that @xmath2 is spin cobordant to zero
. then one can find a metric @xmath91 on @xmath26 such that @xmath254 on the product ends and @xmath1234 in the sobolev @xmath29 completion .
[ r : two ] note that the conditions of this theorem are obviously satisfied for spin 4-manifolds @xmath2 with integral homology of @xmath7 because the spin cobordism class of such a manifold is determined by its @xmath503genus , @xmath1235 the rest of this section will be dedicated to the proof of theorem [ t : two ] .
we showed in the proof of theorem [ t : one ] that @xmath2 can be obtained from @xmath1214 by performing surgery along spheres disjoint from the product region @xmath255
\times y$ ] . in the language of manifolds with product ends
, this implies that @xmath26 can be obtained from the product manifold @xmath508 by performing surgeries inside a compact region in @xmath508 .
since @xmath1212 , the sobolev @xmath29 completion of the operator @xmath1236 is obviously invertible .
therefore , we can proceed with the construction of a desired metric on @xmath26 exactly as in @xcite making some changes along the way to account for the non - compactness of @xmath26 .
the first change comes up in the proof of ( * ? ? ?
* lemma 3.4 ) which uses rellich lemma to conclude that an @xmath1237 bounded sequence of spinors contains a strongly convergent subsequence in @xmath29 .
while rellich lemma fails on non - compact manifolds , the sequence of _ harmonic _ spinors in lemma 3.4 still admits a strongly convergent subsequence : we first apply rellich lemma on the compact manifold @xmath1238\times y)\,\cup\,w\ , \cup\ , ( [ 0,1 ] \times y)$ ] , and then use the following estimate .
[ l ] let @xmath1239 be a manifold with product metric @xmath254 such that @xmath1212
. then there exists a constant @xmath515 such that , for any harmonic @xmath29spinor @xmath1240 on @xmath2 , @xmath1241\times y)}.\ ] ] let @xmath1242 be an orthonormal basis of eigenspinors of @xmath137 and let @xmath1243 stand for the smallest positive @xmath1244 . since @xmath1240 is an @xmath29spinor in the kernel of the operator @xmath1245 , it takes the form @xmath1246 a direct calculation with this formula gives @xmath1247 and @xmath1248\times y)}\ ; = \ ; \int_0 ^ 1\;\|\varphi\|^2_{l^2(y)}\,dt\ ; = \ ; \sum_{\lambda > 0}\ ; \frac { \;\,|a_{\lambda}|^2}{2\lambda}\cdot\left(1 - e^{-2\lambda}\right).\ ] ] this leads to the desired estimate with the constant @xmath1249 .
the rest of the proof of lemma 3.4 goes through using exhaustion of the complement of the surgery sphere @xmath1250 in @xmath26 by the compact sets @xmath1251\times y)\,\cup\,w\ , \cup\ , ( [ 0,1/{\varepsilon } ] \times y ) - u_s ( { \varepsilon}),\ ] ] where @xmath1252 is the open tubular neighborhood of @xmath1250 of radius @xmath256 .
all the lemmas used in that proof are already proved in ( * ? ? ?
* section 2 ) without the compactness assumption .
the second change comes up in the proof of step 2 on page 537 of @xcite .
that step goes through using exhaustion by the compact sets @xmath1253\times y)\,\cup\,w\ , \cup\ , ( [ 0,z ] \times y ) - u_s ( 1/z)\ ] ] for positive integers @xmath110 . in step 3 ,
convergence in @xmath1254 of a sequence of harmonic spinors implies only @xmath29 convergence on compact subsets of @xmath1255 . to obtain the desired @xmath29 convergence on the entire @xmath1255
, we use lemma [ l ] one more time .
let @xmath2 be a connected smooth spin compact manifold of dimension @xmath500 and @xmath736 a smooth map such that the cohomology class @xmath737 \in h^1 ( x ; \mathbb z)$ ] is primitive .
choose a connected manifold @xmath14 poincar dual to @xmath737 $ ] .
we will assume that the manifold @xmath9 with the induced spin structure is a spin boundary and that the @xmath503genus of @xmath2 vanishes ; both of these conditions are automatic when @xmath2 is a homology @xmath7 .
define the riemannian manifold @xmath506 with long neck as in . recall that the metric @xmath505 on @xmath506 takes the form @xmath1257 along the neck .
consider the holomorphic family @xmath1258 under the assumption that @xmath1259 for all @xmath745 on the unit circle @xmath740 , the periodic @xmath1256invariant @xmath1260 was defined in @xcite by the formula it follows from ( * ? ? ? * theorem a and remark 5.4 ) that the so defined periodic @xmath1256invariant is independent of the choice of @xmath1262 as long as @xmath1263 is supported in the product region in @xmath506 , which we will assume from now on .
[ t : eta ] let @xmath1264 be the atiyah patodi
singer @xmath1256invariant @xcite of the dirac operator @xmath1265 . used in this paper and in @xcite . ]
assume that the operator @xmath1266 has zero kernel , as does the @xmath29completion of the operator @xmath513 on the manifold @xmath26 obtained from @xmath15 by attaching infinite product ends .
then the invariants @xmath1260 are well defined and , for all sufficiently large @xmath267 , @xmath1267 the well definedness of @xmath1268 for all sufficiently large @xmath267 follows from proposition [ eigenvalue estimate_z ] . according to ( * ? ? ? * theorem a and remark 5.4 ) , for any spin manifold @xmath507
whose periodic end is modeled on @xmath1269 we have @xmath1270 a similar formula applied to an end - periodic manifold @xmath247 whose end is modeled on @xmath1271 yields @xmath1272 we know from ( * ? ? ?
* section 6.3 ) that @xmath1273 . by subtracting the above formulas from each other ,
we conclude that @xmath1260 must differ from @xmath1274 by an even integer .
the statement of the theorem will follow as soon as we prove that @xmath1260 and @xmath1275 can be made arbitrarily close by choosing sufficiently large @xmath267 .
the proof of this will occupy the rest of this section . because of the periodic index theorem @xcite the statements of theorem [ thm ] and theorem [ t : eta ] are essentially equivalent to each other .
what follows is an independent proof of theorem [ t : eta ] using heat kernel techniques .
we chose to include this proof , inspired by @xcite , because it may be of interest in its own right .
let @xmath1277 a circle of length @xmath562 and consider a smooth function @xmath1262 on its universal cover , the real line , such that @xmath1278 .
the cohomology class of @xmath1263 generates @xmath1279 , and we will write @xmath1280 with respect to the natural parameter @xmath1281 on the circle @xmath1277 . consider the family of elliptic operators @xmath1282 with @xmath1283 .
if @xmath1284 is the kernel of the operator @xmath1285 then @xmath1286 is the kernel of the operator @xmath1287 [ l : taubes ] there are positive constants @xmath56 and @xmath515 independent of @xmath745 and @xmath562 such that the following estimates hold for all unitary @xmath745 , @xmath1288 , and @xmath1289 : @xmath1290 we begin by observing that , for the purpose of making kernel estimates on the circle @xmath740 , one may assume that @xmath1291 is constant and is therefore equal to @xmath1292 .
this can be seen from the formula @xmath1293 which holds for any function @xmath48 defined on @xmath1277 , and the fact that adding @xmath1294 does not change the cohomology class of @xmath1263 .
choosing @xmath1295 will then do the job .
conjugating the operator by the unitary complex number @xmath1296 multiplies the kernel by @xmath1297 hence preserves its norm .
let @xmath1298 then the kernel @xmath1284 of @xmath1299 is given by the formula @xmath1300 with respect to the orthonormal basis @xmath1301 on the circle @xmath1277 .
one can easily verify that @xmath1302 where @xmath1303 stands for the heat kernel of @xmath1304 on the circle of length one .
it is well known that there exist positive constants @xmath56 and @xmath626 independent of @xmath745 such that @xmath1305 but then latexmath:[\[\begin{gathered } \notag given a closed spin riemannian manifold @xmath9 of dimension @xmath1308 , consider the chiral spin dirac operators @xmath1309 on @xmath1310 and their twisted versions @xmath1311 .
we wish to derive estimates on the kernels of the operators @xmath1312 and @xmath1313 which are uniform in @xmath745 on the unit circle @xmath740 and in @xmath562 .
denote by @xmath1314 the kernel of the operator @xmath1312 then @xmath1315 is the kernel of @xmath1316 .
[ l : est1 ] suppose that the dirac operator @xmath1266 on @xmath9 has zero kernel .
then there exist positive constants @xmath56 and @xmath626 independent of @xmath745 and @xmath562 such that the following estimates hold for all unitary @xmath745 , @xmath1288 , and @xmath1289 : latexmath:[\[\begin{gathered } on the circle @xmath740 , we have @xmath1318 and @xmath1319 hence @xmath1320 and the kernel of @xmath1321 is the product of the kernels of @xmath1285 and @xmath1322 . to obtain estimate , simply combine the estimate @xmath1323 on the former kernel with the estimate @xmath1324 for @xmath1325 on the latter ; see lemma [ l : taubes ] and ( * ?
* proposition 1.1 ) , respectively . to obtain estimate ,
write the operator @xmath1326 in the form @xmath1327\cdot \exp(-t { \mathcal d}^2 ) \hspace{0.4 in } \\ - du\cdot\exp\,(t\cdot ( { \partial}/{\partial}u - \ln z\cdot f')^2)\cdot \left[{\mathcal d}\exp(-t{\mathcal d}^2)\right ] \end{split}\ ] ] and apply the estimates of lemma [ l : taubes ] and ( * ? ? ?
* proposition 1.1 ) twice .
[ l : est0 ] suppose that the dirac operator @xmath1266 on @xmath9 has zero kernel .
then there are positive constants @xmath1328 and @xmath626 independent of @xmath745 and @xmath562 such that the following estimates hold for all unitary @xmath745 , @xmath1288 , and @xmath1329 : @xmath1330 we use again the fact that the kernel of @xmath1321 is the product of the kernels of @xmath1331 and @xmath1322 .
the former is uniformly bounded for all @xmath1329 by lemma [ l : taubes ] . as for the latter ,
note that the smallest eigenvalue of @xmath1332 is positive .
denote this eigenvalue by @xmath660 then the kernel of @xmath1333 can be estimated from above by @xmath1334 by ( * ? ?
* proposition 1.1 ) .
the argument for the kernel of @xmath1335 is similar using equation together with ( * ? ? ?
* proposition 1.1 ) .
let @xmath9 be a connected submanifold of @xmath2 which is poincar dual to @xmath1263 , and let @xmath15 be the cobordism obtained by cutting @xmath2 open along @xmath9 .
assume that the riemannian metric on @xmath2 is a product metric in a normal neighborhood @xmath1336 \times y$ ] .
the induced metric on @xmath15 will have product regions @xmath1337 \times y$ ] and @xmath851 \times y$ ] near its boundary components .
we will use these product regions to define , for every real number @xmath1338 , the manifold @xmath1339
\times y)\ ] ] by gluing the product region @xmath1337 \times y$ ] of the first summand to @xmath1340 \times y$ ] of the second , and the product region @xmath851 \times y$ ] of the first summand to @xmath1341 \times y$ ] of the second .
the gluing functions we use are linear on the first factor and are the identity on the second .
we will view the manifold @xmath1342
\times y$ ] with the identified boundary components as the product @xmath1343 with the circle @xmath1344 of circumference @xmath1345 , cut open along a copy of @xmath9 . throughout this section
we assume that the dirac operator @xmath1266 on @xmath9 has zero kernel , @xmath1263 is supported in the product region @xmath1336 \times y$ ] , and @xmath1262 is a function of the normal coordinate @xmath1281 in that region .
denote by @xmath1314 the kernel of the operator @xmath1346 on @xmath506 and by @xmath1347 and @xmath1348 the kernels of the operators @xmath1346 on , respectively , @xmath2 and @xmath1343 .
we wish to compare the functions @xmath1349 and @xmath1350 over the product region @xmath1336
\times y$ ] shared by the manifolds @xmath506 and @xmath1343 . to this end , define an approximate kernel of the operator @xmath1351 on @xmath506 by the formula @xmath1352 the functions @xmath1353 and @xmath1354 here form a smooth partition of unity on @xmath506 such that @xmath1355 \times y$ ] and @xmath1356 on @xmath1357 \times y$ ] .
the function @xmath1358 equals zero on @xmath1359 \times y$ ] and one outside of @xmath1360\times y$ ] , and the function @xmath1361 equals one on @xmath1362 \times y$ ] and zero outside of @xmath1363\times y$ ] .
note that @xmath1364 on @xmath1365 and that the distance between @xmath1366 and @xmath1365 is greater than or equal to @xmath1367 , @xmath1368 .
the advantage of having the approximate smoothing kernel @xmath1369 is that it is defined on the same manifold @xmath506 as @xmath1349 while @xmath1370 in the region @xmath1336
\times y$ ] of our interest . to calculate the latter difference ,
consider the error term @xmath1371 where the operator @xmath1372 acts on the variable @xmath452 for any fixed @xmath1062 and @xmath1373 .
since @xmath1374 and both @xmath1375 solve the heat equation , we obtain @xmath1376 the @xmath1380integration in this formula extends only to @xmath1381 , which is contained in @xmath1382 \times y)\ ; \cup\ ; ( [ r + 1/7,r+6/7 ] \times y)$ ] .
therefore , the above integral can be written in the form @xmath1383 [ p : est1 ] there are positive constants @xmath18 , @xmath56 , and @xmath626 independent of unitary @xmath745 and @xmath267 such that the following estimate holds for @xmath1389 \times y \subset x_r$ ] and @xmath1289 : @xmath1390 [ l : est2 ] there are positive constants @xmath18 , @xmath56 , and @xmath626 independent of unitary @xmath745 and @xmath267 such that the following estimates hold for @xmath1392 and @xmath1289 : @xmath1393 , such estimates hold on @xmath1343 with @xmath1394 and the constants @xmath56 and @xmath626 independent of @xmath745 and @xmath267 . according to ( * ? ? ?
* proposition 1.1 ) , the same estimates with @xmath1394 hold on the manifold @xmath2 but only for non - twisted dirac operators , which are exactly the operators that @xmath2 contributes into the definition of the approximate kernel .
now , the kernels of @xmath1395 and @xmath1326 on @xmath506 can be constructed from this approximate kernel by an iterative procedure using the duhamel principle as in the proof of ( * ? ? ?
* theorem 2.4 ) . in the process
, one obtains the estimates and on @xmath506 from the respective estimates on @xmath1343 and @xmath2 .
the constants in these estimates will be independent of @xmath745 and @xmath267 because they were already independent of @xmath745 and @xmath267 on @xmath1396 and @xmath2 .
one also acquires in the process a possibly non - zero constant @xmath18 which has to do with the volume of @xmath9 and is therefore independent of @xmath745 and @xmath267 .
there exist positive constants @xmath56 and @xmath626 independent of unitary @xmath745 and @xmath267 such that the following estimate holds for @xmath1392 and @xmath1289 : @xmath1397 and @xmath1348 and their space derivatives . that the estimates for @xmath1348 are independent of @xmath745 and @xmath267 follows as in the proof of lemma [ l : taubes ] from an explicit formula for the heat kernel on @xmath1343 .
the negative powers of @xmath1062 that show up in the estimates are absorbed into the factor @xmath1398 using the observation that @xmath1377 whenever @xmath1378 .
we can now proceed with estimating the difference @xmath1399 for @xmath1389 \times y$ ] using formula . for all
@xmath1400 we have @xmath1401 hence @xmath1402 and gives @xmath1403 similarly , using , we obtain @xmath1404 hence @xmath1405 the second estimate that goes into the proof of our theorem has to do with the smallest eigenvalue of the operator @xmath1266 on @xmath9 . such estimates are well - known for the @xmath29norms of heat kernels ; the following proposition claims pointwise estimates .
we begin with an observation about the sobolev spaces on @xmath506 with a fixed @xmath267 . for a unitary @xmath745 and a non - negative integer @xmath413 , define the sobolev norm @xmath1410 this norm is equivalent to the sobolev norm @xmath1411 meaning that the identity operators @xmath1412 and @xmath1413 are bounded .
the norms of these operators are continuous functions of @xmath745 which achieve their absolute minimum and maximum on the circle @xmath740 .
therefore , there exist positive constants @xmath483 and @xmath303 independent of @xmath1414 such that for all @xmath98 we have @xmath1415 with this understood , let @xmath1416 be the spectral decomposition of the full dirac operator @xmath1417 with @xmath1418 ( both @xmath1419 and @xmath1420 depend on @xmath745 but we omit this dependence from our notation ) .
we first estimate @xmath1421 use the sobolev embedding theorem with @xmath1422 , see for instance ( * ? ? ?
* lemma 1.1.4 ) , and inequality to obtain the following pointwise estimates @xmath1423 with the constants @xmath467 and @xmath483 independent of @xmath745 and @xmath1424 . for any @xmath1329
, we have @xmath1425 with some positive constants @xmath1328 and @xmath1426 independent of @xmath745 . in the last line
, we used the condition that @xmath1427 on the circle @xmath740 to guarantee that the smallest eigenvalue of the family @xmath1428 over @xmath1414 is positive ; we call this eigenvalue @xmath660 .
we also used the fact that @xmath1429 is a continuous function of @xmath1414 to estimate it from above by a constant independent of @xmath745 .
the constants in the estimates depend on @xmath267 in two different ways .
one is via the smallest eigenvalue @xmath1244 which by proposition [ eigenvalue estimate_z ] is bounded away from zero by a positive constant for all sufficiently large @xmath267 .
the other is via the function which can be shown to be bounded for @xmath1329 using ( * ? ?
* lemma 10.13 ) and proposition [ eigenvalue estimate_z ] . for our choice of @xmath1263 ,
the difference between @xmath1260 and @xmath1430 is given by integrating the quantity @xmath1431\times y}\ ; { \operatorname{tr}}\ , ( df\cdot ( { \mathcal d}^+_z k_z ( t;x , x ) - { \mathcal d}^+_z k^a_z(t;x , x)))\,dx,\ ] ] with respect to @xmath1062 and @xmath745 as in formula ( in the above formula , @xmath1432 stands for the matrix trace ) . the quantity @xmath1433 has been estimated twice , first by @xmath1434 for @xmath1289 in proposition [ p : est1 ] and then by @xmath1435 for @xmath1329 in proposition [ p : est3 ] .
the graphs of @xmath1436 and @xmath1437 intersect at the point @xmath1438 , where @xmath1439 therefore , for sufficiently large @xmath267 and after adjusting the constants @xmath626 , we can use the first estimate on the interval @xmath1440 and the second on the interval @xmath1441 .
since both estimates are independent of @xmath452 and @xmath745 , we can integrate them with respect to these two variables .
the integration results in estimating the distance between @xmath1260 and @xmath1430 by a uniform constant times the integral the latter integral can easily be estimated from above by @xmath1443 since @xmath1444 is negative , the difference between @xmath1260 and @xmath1430 must approach zero as @xmath28 .
this completes the proof of theorem [ t : eta ] .
height 2pt depth -1.6pt width 23pt , _ the equivalence of heegaard floer homology and embedded contact homology via open book decompositions iii : from hat to plus_. preprint http://arxiv.org/abs/1208.1526 , 2012 .
height 2pt depth -1.6pt width 23pt , _ an introduction to knot floer homology _ , in `` proceedings of the 2013 sms summer school on homology theories of knots and links '' .
preprint https://arxiv.org/abs/1401.7107 , 2014 .
c. manolescu and p. ozsvth , _ on the khovanov and knot floer homologies of quasi - alternating links _ , in `` proceedings of gkova geometry - topology conference 2007 '' , gkova geometry / topology conference ( ggt ) , gkova , 2008 , 6081 .
height 2pt depth -1.6pt width 23pt , _ casson type invariants in dimension four _ , in `` geometry and topology of manifold '' , fields institute communications 47 , amer .
soc . , providence , ( 2005 ) , 281306 . | we study the seiberg witten invariant @xmath0 of smooth spin @xmath1-manifolds @xmath2 with rational homology of @xmath3 defined by mrowka , ruberman , and saveliev as a signed count of irreducible monopoles amended by an index - theoretic correction term .
we prove a splitting formula for this invariant in terms of the fryshov invariant @xmath4 and a certain lefschetz number in the reduced monopole floer homology of kronheimer and mrowka .
we apply this formula to obstruct existence of metrics of positive scalar curvature on certain 4-manifolds , and to exhibit new classes of homology @xmath5-spheres of infinite order in the homology cobordism group . |
cataclysmic variables ( cvs ) are semi - detached binaries with a white dwarf ( wd ) primary and a low - mass companion which transfers mass to a wd through roche - lobe overflow .
the observed orbital period distribution of cvs has three important features : a sharp short period cut - off at the orbital period of @xmath10 min , referred to as the `` minimum period '' , the period gap , a dearth of systems with periods between @xmath11 min and @xmath12 min @xcite , and the `` period minimum spike '' , a significant accumulation of systems at @xmath13@xmath14 min @xcite . for cvs with period above @xmath15 ,
the dominant angular momentum loss mechanism is thought to be the magnetic braking , while the gravitational radiation may be responsible for the angular momentum loss for cvs below @xmath16 . in this work ,
we focus on cvs only with @xmath17 .
a characteristic feature of cvs is the mass transfer from a secondary to a wd .
since the gas stream carries substantial angular momentum , it forms an accretion disk around the wd .
the reaction of a cv on the mass transfer can be qualitatively understood with the help of the dimensionless parameter @xmath18 where @xmath19 and @xmath20 denote the mass - loss and kelvin - helmholtz timescales of the donor , respectively @xcite . at the initial phase of mass transfer , @xmath21 , so that the donor remains close to thermal equilibrium and thus follows the mass - radius relation ( mrr ) of main - sequence stars @xmath22 with the effective mass - radius exponent @xmath23 close to @xmath24 @xcite . here
, the subscript `` 2 '' denotes the quantities of cv donors . both @xmath25 and @xmath1 of the cv decrease as the donor keeps losing mass during this early phase of cv evolution .
as the mass transfer continues , the binary shrinks due to the angular momentum loss , reducing @xmath26 . when @xmath19 and @xmath20 become comparable to each other ( @xmath27 )
, the donor becomes out of thermal equilibrium .
since the surface luminosity is no longer in balance with the nuclear luminosity , the donor expands , causing @xmath23 to decrease with time @xcite . using the hayashi theory and homologous approximation ,
@xcite studied the reaction of the donor to mass loss and found that @xmath25 increases with decreasing @xmath26 .
observations confirm this theoretical prediction , showing that the radii of donors in short - period cvs are @xmath28@xmath29 larger than those of isolated main - sequence stars @xcite .
the effective mass - radius exponent @xmath23 becomes smaller as @xmath30 decreases towards the brawn dwarfs masses .
the change in the mrr of donors slows down the temporal change in the orbital period as the donor continues losing mass .
the orbital period reaches a minimum value when @xmath31 at which @xmath32 @xcite . as @xmath26 decreases further , @xmath23 declines to @xmath33 @xcite .
this in turn leads to an increase in @xmath1 .
such a turnaround of the orbital period at @xmath34 is referred to as the `` period bounce '' .
the evolutionary scenario described above suggests that the mass - loss rates @xmath2 ( and the related thermal relaxation ) of donors control the evolution of cvs .
it also implies that the observed distribution of cvs versus the orbital period may be a consequence of a certain relation between @xmath2 and @xmath1 .
since @xmath35 not only provides a way to estimate the ages of cvs but also can be used to check the dominant mechanism for angular momentum loss , finding a reliable @xmath1@xmath2 relationship is of great importance in understanding cv evolution .
there have been numerous studies to derive the @xmath1@xmath2 relation ; these can be categorized into two groups : purely theoretical work ( e.g. , @xcite ) and semi - empirical work ( e.g. , @xcite ) . the first , theoretical studies made use of stellar evolution codes combined with the analytical mass - transfer models of @xcite or others .
the common results of these numerical experiments are that ( 1 ) the orbital period exhibits the turnaround behavior at @xmath34 that is about @xmath36 smaller than the observed @xmath37 @xcite and ( 2 ) compared to isolated stars , the secondaries in short - period cvs are oversized due to the mass loss but only by @xmath28@xmath38 , which is about a factor @xmath39@xmath40 smaller than the observed cv sizes .
since the bloating of the donor depends on the efficiency of the mass loss , these results suggest either the numerical models underestimate @xmath2 above @xmath34 , or the efficiency of the angular momentum loss is higher than expected @xcite . despite the attempts to include the effects of tidal and rotational perturbations @xcite , additional angular momentum loss via magnetic braking @xcite , accretion disk winds and magnetic propeller ( see @xcite for review ) , etc . , clear explanations for the discrepancies between @xmath34 and @xmath37 and between the empirical and numerical mrrs have yet to be found .
the second , semi - empirical approach uses , for instance , the effective temperatures of wds @xcite or accretion disk luminosities @xcite to infer the accretion rates @xmath41 .
the physical basis for the former is that wds in cvs are observed to be hotter than expected from their ages @xcite , and such overheating " of wds may be a result of the compressional heating due to the gas piled up on their surfaces @xcite . for the latter method ,
the accretion luminosities can give direct information on @xmath41 if the disk properties are well constrained . when applied to dwarf novae ( dne ) that accrete sporadically and have thermally unstable disks , @xmath41 calculated from the two methods have a different scatter for given @xmath1 .
the mass - transfer rates based on the accretion luminosities probably measure an accretion rate averaged over short time intervals , e.g. , a few decades , which makes them quite sensitive to short - time disk variability @xcite .
this inevitably leads to a high dispersion in @xmath41 estimated from the disk luminosities . on the other hand ,
the effective temperatures of wds are likely to give a relatively long - term averaged @xmath41 , although this method may suffer from uncertainties related to unknown properties such as masses and radii of wds as well as some auxiliary assumptions on their structures and boundary layers @xcite . in this paper
we construct another semi - empirical relation between the mass - loss rate and the orbital period , and compare the result with those of the purely theoretical and the semi - empirical approaches mentioned above .
we start from the empirical mrr of @xcite for superhumping cvs , and apply the homologous approximation in order to calculate the dependencies of the effective temperature and gravothermal luminosity upon the bloating of the donor . assuming that the observed mrr of cv donors is a consequence of the mass loss and thermal relaxation processes , we calculate the mass - loss rates .
we will show that the resulting mass - loss rates are broadly in line with those on the basis of the effective temperatures of wds , while different from those of the purely theoretical models of @xcite .
we will also show that our results are consistent with the angular momentum loss of cvs due primarily to the emission of gravitational waves . since radii of cv donors likely trace the long - term evolution of @xmath2 @xcite and depend only weakly on details of the mass - transfer processes , the resulting mass - loss rates from the empirical mrr are less model - dependent and less sensitive to the short - term effects than the other semi - empirical estimates that rely on accretion activities .
the organization of this paper is as follows . in
[ sec_p_te ] , we introduce the theoretical mrr for isolated single stars and the empirical mrr for cvs donors , and define the bloating factor .
we then describe a method to find the effective temperature and gravothermal luminosity of donors by using the bloating factor . in
[ sec_dm ] , we calculate the semi - empirical mass - loss rate , construct the mass - loss history to estimate the cv ages , and show that the derived mass - loss rate is consistent with cv evolution by angular momentum loss via the emission of gravitational waves . in
[ sec_dis ] , we apply our results to selected eclipsing short - period cvs and show that the semi - empirical mass - loss rate is consistent with the accretion rates inferred from the effective temperatures of wds .
we also suggest a way to estimate the effective temperatures of donors assuming that mass is conserved .
finally , we summarize our results in [ sec_con ] .
we consider a cv consisting of a wd ( or accretor ) with mass @xmath42 and a secondary ( or donor ) with mass @xmath30 , radius @xmath25 , and effective temperature @xmath0 .
we assume that the donor fills its critical equipotential surface . in accordance with the results of @xcite and @xcite
, we assume @xmath43 throughout this paper .
we use an additional subscript `` 0 '' to denote the quantities of isolated , single stars in hydrostatic and thermal equilibrium to which cv secondaries will be compared . in this section
, we use the empirical relationship between masses and radii of cv secondaries to calculate the bloating factor in comparison with the single stars of the same mass .
we then treat the bloating factor as perturbations to the properties of single stars in order to find the orbital period , effective temperature , and gravothermal luminosity of cvs . as the mrrs of single stars that we use as the references to those of cv secondaries
, we take _ composite _ relations constructed in a manner similar to @xcite : the bcah98 isochrone of @xcite for stars with @xmath44 , the dusty isochrone of @xcite for @xmath45 ( @xmath46 k ) , and the cond isochrone of @xcite for @xmath47 .
there are quite large uncertainties in the estimation of cv ages .
period bouncers are thought of as a final stage of cv evolution , suggesting that they are older than their `` parent '' cvs ( see , e.g. , @xcite ) .
although numerical experiments predict that cvs cross the period gap at @xmath48 gyr and reach @xmath34 in @xmath49 gyr after the onset of the mass loss below the gap ( e.g. , @xcite ) , temporal evolution of cvs depends strongly on the mass - loss rate that is not well constrained observationally . to cover a wide range of cv ages ,
we adopt two composite isochrones with 1-gyr and 5-gyr ages ( hereafter b1 and b5 sequences , respectively ) and 1-gyr isochrone for @xmath50 . the choice of the 1-gy isochrone in @xcite was to represent the numerical results of @xcite . ] .
figure [ pic_m_r ] plots the composite @xmath51 of single stars as functions of @xmath30 for the b1 and b5 sequences , as dashed and solid lines , respectively .
both sequences give almost the same radii above @xmath52 , while below this point the stellar radii in the b1 sequence are about @xmath36 larger than in the b5 sequence . as the mrr of cv secondaries ,
we take the relation derived by @xcite from the observed properties of superhumping cvs : @xmath53 this empirical fit assumes @xmath43 and incorporates itself the additional empirical constraints that reproduce the observed locations of the period gap and the period minimum , such that @xmath16 occurs at @xmath54 and @xmath37 at @xmath55 .
the intrinsic errors in @xmath25 for this empirical mrr are estimated to be about @xmath40@xmath56 @xcite ; we take allowance for @xmath57 uncertainties in @xmath25 throughout this work .
we note a caveat that our choice of the errors may underestimate real values below @xmath52 where only a few systems are observed and thus there is practically no calibrator .
in addition , the masses of wds may differ from the assumed mass of @xmath58 , which also can change @xmath25 .
equation ( [ eq_kn_mr_fit ] ) is plotted in figure [ pic_m_r ] as dotted line with the shaded region denoting the associated uncertainties in @xmath25 .
columns ( 1 ) and ( 2 ) in table [ ta_seq ] give the @xmath30-@xmath25 relation .
for comparison , figure [ pic_m_r ] also plots as a dot - dashed line the mrr of cvs with @xmath59 from the numerical models of @xcite , which clearly has smaller radii than the empirical relation , typically by @xmath28@xmath60 .
this indicates that the deviation of real donors from thermal equilibrium is larger than that expected from the numerical models .
we define the bloating factor @xmath61 as the relative size of a cv secondary to a single star with the same mass .
figure [ pic_m_al ] plots @xmath62 as dashed and solid lines for the b1 and b5 sequences , respectively , with the shaded and hatched regions representing the corresponding uncertainties .
the discontinuity of @xmath62 at @xmath63 corresponds to the period gap . since @xmath64 ( see below ) , the ratio of the bloating factors above ( @xmath65 ) and below ( @xmath66 ) the gap satisfies @xmath67 , in good agreement with the width of the gap . at the upper edge of the gap ,
the donors detach from their critical lobes and the mass loss stops .
since the donors are thought to be oversized due to the mass loss , they shrink within the gap . when the donors come out the lower edge of the gap , they resume the mass loss and become oversized again .
figure [ pic_m_al ] shows that the bloating factor increases with the decreasing donor mass for @xmath68 and then decreases with decreasing @xmath30 below this point .
this can be explained qualitatively as follows .
since stars with @xmath69 are fully convective and isentropic , they are well described by polytropes with index @xmath70 @xcite .
the mrr for these polytropes @xmath71 with @xmath72 being the pressure constant @xcite , then gives @xmath73 the mass loss is capable of driving the donor out of thermal equilibrium .
this raises the gravothermal luminosity , @xmath74 , where @xmath75 is the nuclear luminosity of the donor and @xmath76 is the surface luminosity . for fully convective stars ,
the change of @xmath72 depends on @xmath77 and @xmath20 as @xmath78 @xcite . combining equations
( [ eq_r_m_polytrope ] ) , ( [ eq_alpha_def_k ] ) , and ( [ eq_dk_k ] ) , one obtains @xmath79 where @xmath80 is the mass - radius exponent of unperturbed , singe stars . at the lower edge of the period gap , a donor remains close to thermal equilibrium so that @xmath81 and @xmath82 . in this case , the second term in the right - hand side of equation ( [ eq_da_a ] ) dominates and @xmath83 , indicating that @xmath62 increases as the mass loss continues .
when the donor reaches the minimum hydrogen burning mass ( mhbm ) @xmath84 , the nuclear burning stops and donors are born as brown dwarfs , so that @xmath85 . in the brown dwarf regime , on the other hand , the internal pressure is dominated by the degeneracy pressure with @xmath72 depending mainly on the chemical composition @xcite . in this case , @xmath80 tends to @xmath86 and the first term in the right - hand side of equation ( [ eq_da_a ] ) dominates , resulting in @xmath87 . for brown - dwarf donors ,
the change in the equation of state partially compensates for the increasing tendency of the bloating factor due to the mass loss .
in addition to the mass loss , tidal and rotational perturbations can also change the donor size , depending on the polytropic index and the mass ratio .
we assume that the applied perturbations do not change the specific entropy , which is reasonable since the stellar distortions occur in dynamical timescales . in appendix
[ ap_isentropic ] , we calculate the enlargement factor @xmath88 of a roche - lobe filling donor as a function of @xmath89 for @xmath90 .
the resulting @xmath88 for @xmath70 polytropes against @xmath30 is plotted in figure [ pic_m_al ] as a dotted line .
note that the expansion of low - mass donors due to the tidal and rotational distortions is @xmath49@xmath91 , insensitive to the mass ratio for @xmath92@xmath93 ( see also , e.g. , @xcite ) .
it can be seen that while @xmath88 is , in general , considerably smaller than the observed bloating factor for stars with @xmath94 , it is comparable to @xmath62 when @xmath95 corresponding to @xmath16 or when @xmath96 as donors approach @xmath37 .
this indicates that the effect of the isentropic expansion of donors on the bloating factor is non - negligible for cvs that just resume mass loss or are below the minimum period . for a given mrr
, the orbital period can be determined by the roche lobe geometry and kepler s law as @xmath97 where @xmath98 is the mean density of the donor , @xmath99 is the relative size of the donor to the orbital separation , and @xmath100 is the mass ratio ( e.g. , @xcite ) . equation ( [ eq_mean_den_law ] ) implies that the orbital period depends on the bloating of the donor as @xmath101 clearly , the orbital period increases with the radius of a secondary .
since donors below the period gap are fully convective , their luminosities are determined by the conditions at the very outermost layers @xcite , so that one can neglect the influence of the distortions on the central parts of a donor . to obtain the properties of the photospheric layers , we use the standard hayashi theory . for the opacity law of the form
@xmath102 with the gas pressure @xmath103 , the ratio of the effective temperatures satisfies @xmath104 where @xmath105 ( e.g , @xcite , see also @xcite ) . for the realistic opacity laws applicable for low - mass main - sequence stars
, @xmath106 has a very small value .
if the opacity is governed by h@xmath107 , for example , @xmath108 and @xmath109 , leading to @xmath110 . if the molecular absorption instead is a dominant opacity source ( @xmath111 and @xmath112 ) , @xmath113 .
therefore , donors in short - period cvs have approximately the same effective temperatures as the corresponding isolated single stars of identical masses ( e.g. , @xcite ) .
to construct the semi - empirical @xmath1@xmath0 relations , we first use the theoretical @xmath30@xmath114 relation for isolated stars and calculate @xmath115 from equation ( [ eq_mean_den_law ] ) by fixing @xmath43 .
we then use the empirical mrr and the bloating factor of donors to calculate @xmath1 and @xmath0 from equations ( [ eq_p_cor ] ) and ( [ eq_te_cor_kw ] ) .
the @xmath115@xmath114 relations are plotted in figure [ pic_te_p ] as thin lines , while the thick lines are for the semi - empirical @xmath1@xmath0 relations .
the dashed and solid lines correspond to the b1 and b5 sequences , respectively . again , the shaded and hatched regions give the corresponding uncertainties .
columns ( 3 ) , ( 6 ) , and ( 11 ) in table [ ta_seq ] list the values of @xmath1 and @xmath0 against @xmath30 for both sequences . for a given mass , the expansion of a donor increases the orbital period while the effective temperature keeps almost unchanged .
the largest difference between @xmath1 and @xmath115 occurs at @xmath116 k , corresponding to @xmath117 where the bloating factor is maximal .
the dot - dashed line plots the semi - empirical result of @xcite . because of the difference in the adopted ages of the isochrones , the result of @xcite for @xmath68 ( with @xmath118 k ) and @xmath47 ( with @xmath119 k ) is in agreement with our results for the b5 and b1 sequences , respectively .
the donors at @xmath37 have mass @xmath120 , corresponding to @xmath121@xmath122 k and @xmath123@xmath124 k for the b5 and b1 sequences , respectively . the right choice of the stellar age is quite important in evaluating the mass - loss rate near @xmath37 since it depends rather sensitively on @xmath0 as @xmath125 @xcite .
combining equation ( [ eq_te_cor_kw ] ) with the stefan - boltzmann law @xmath126 where @xmath127 is stefan - boltzmann constant , one can show that the change in the stellar luminosity is given by @xmath128 although the effective temperature does not change much with the bloating factor , the additional @xmath129 term in the stefan - boltzmann law increases @xmath130 significantly .
columns ( 5 ) and ( 10 ) in table [ ta_seq ] give the values of @xmath76 for the b1 and b5 sequences , respectively .
the nuclear luminosity @xmath75 is determined by the energy produced in the central parts of a donor , so that @xmath131 , where @xmath132 is the thermonuclear reaction rate given by @xmath133 with @xmath134 and @xmath135 denoting the central temperature and density , respectively . since @xmath136 for a monatomic ideal gas ,
the change in the nuclear luminosity due to the bloating factor is @xmath137 theoretical models for stellar structure and evolution give @xmath138 and @xmath139 @xcite for @xmath140 k and @xmath141
g @xmath142 , typical for the centers of low - mass stars . since @xmath143 for isolated stars in thermal equilibrium , the gravothermal luminosity satisfies @xmath144 thermonuclear burning stops in cv secondaries once they reach the mhbm @xmath84 . at this point
, donors are effectively born as brown dwarfs and the gravitational collapse becomes the dominant source of the stellar radiation .
that is , @xmath145 for @xmath146 .
figure [ pic_m_lg ] plots @xmath147 versus @xmath30 for the b1 and b5 sequences as dashed and solid lines , respectively , with the uncertainties denoted by the shaded and hatched regions .
figures [ pic_m_al ] and [ pic_m_lg ] together imply that at any given mass the contribution of the gravothermal luminosity increases with the bloating factor .
if @xmath25 increases by @xmath38 , more than @xmath148 of the donor luminosity is generated by the release of thermal and gravitational energies .
finding the theoretical mass - loss rates of cv donors is a daunting task since it involves many complicated processes including thermal and adiabatic relaxations , nuclear evolution , effects of tidal and rotational distortions , orbital angular momentum loss , etc .
, all of which may affect the internal structure significantly .
although the stationary model of @xcite has been used in various theoretical models of cv evolution , it relies on the parameters describing the donor atmosphere in the vicinity of the inner lagrangian point , which are poorly known . by
relating the mass - loss rate with angular momentum loss , on the other hand , the ritter model allows to check whether a proposed angular momentum loss mechanism is indeed responsible for cv evolution . as mentioned in
[ sec : intro ] , the numerical calculations that incorporate the stellar evolution and the mass - loss model of @xcite produced mrrs that differ from the empirical mrr . instead of employing the ritter model to calculate the mass - loss rate , we seek the @xmath2 that is required to reproduce the empirical mrr .
we make use of the well - known relation between the effective mass - radius exponent @xmath23 of the donor and the mass - loss timescale : @xmath149 @xcite , where @xmath150 is the adiabatic mass - radius exponent for fully convective stars .
equation ( [ eq_ze_def ] ) dictates how the mrr should behave in response to the mass - loss and thermal relaxation processes
. when the mass transfer just commences , @xmath151 and the donor response is adiabatic , leading to @xmath152 .
since the ensuing deviation from the adiabatic evolution depends on how the donor reacts thermally on the mass loss , @xmath23 increases linearly with @xmath26 @xcite . combining equations ( [ eq_mean_den_law ] ) , ( [ eq_st_bl ] ) and ( [ eq_ze_def ] ) with the definitions of the kelvin - helmholtz timescale ( @xmath153 ) ) and mass - loss timescale ( @xmath154 ) , one obtains @xmath155 when @xmath156 corresponding to the period bounce , equation ( [ eq_dm_def ] ) recovers equation ( 34 ) of @xcite for the mass - loss rate at the minimum period .
in figure [ pic_dm_p ] we plot the semi - empirical @xmath1@xmath2 relations calculated from ( [ eq_dm_def ] ) for the 1-gyr and 5-gyr isochrones as dashed and solid lines in red color , respectively .
the corresponding uncertainties are given by the shaded and hatched regions .
columns ( 7 ) and ( 12 ) of table [ ta_seq ] list the values of @xmath2 . the semi - empirical mass - loss rate is essentially non - stationary . for @xmath157 min @xmath158 min ,
@xmath2 is in the range of @xmath159 @xmath160 and decreases relatively slowly as @xmath161 .
below 90 minutes , @xmath2 sharply declines towards @xmath37 as a result of a strong dependence of @xmath2 upon @xmath0 , emulating the @xmath1@xmath0 relation shown in figure [ pic_te_p ] .
since the accretion rate decreases with @xmath2 , this rapid decline of @xmath2 is consistent with the observed low accretion activities of cvs within the period spike identified in the sdss samples @xcite .
note that the mass - loss rates for the b1 and b5 sequence are almost similar to each other for masses above the mhbm , suggesting that these are perhaps permanent values . in the brown dwarf regime ,
on the other hand , the ambiguity in the adopted stellar age leads to a large difference in the mass - loss rate , suggesting that it is necessary to use the appropriate isochrones for an accurate estimation of @xmath2 for donors with @xmath162 . for comparison ,
figure [ pic_dm_p ] also plots as a dot - dashed line the @xmath1@xmath2 relation from the numerical models of @xcite , which shows a pronounced `` boomerang '' shape such that @xmath2 is nearly stationary above @xmath34 with @xmath163 .
it also has @xmath2 significantly smaller for @xmath68 than our semi - empirical results . to see what causes the difference between our semi - empirical @xmath1@xmath2 relation and the numerical result of @xcite
, we calculate @xmath2 from equation ( [ eq_dm_def ] ) by using the mrr of @xcite together with the 5-gyr isochrone .
the resulting @xmath1@xmath2 relation is plotted in figure [ pic_dm_p ] as a dotted line , which is overall in good agreement with the numerical results of @xcite .
this suggests that smaller radii of donors in their work are a primary reason for the discrepancies between the results of the current work and @xcite .
this also proves to some extent that our approach provides a reasonable way to estimate the mass - loss rate .
from the semi - empirical @xmath2 , we calculate @xmath26 for the b1 and b5 sequences , which are listed in columns ( 8) and ( 13 ) of table [ ta_seq ] , respectively .
note that donors in cvs below @xmath16 are out of thermal equilibrium with @xmath164 , which is the main cause of the donor expansion .
equations ( [ eq_lg_def ] ) and ( [ eq_dm_def ] ) imply that for given @xmath30 and @xmath23 , the mass - loss rate increases more than @xmath165 due mainly to changes in the orbital period and gravothermal luminosity .
the resulting semi - empirical @xmath1@xmath2 relation still has a boomerang shape , which is much less dramatic than that of @xcite due to the donor bloating and the associated thermal relaxation .
nevertheless , it is very difficult to detect the boomerang shape observationally because of the sharp decline of @xmath0 and @xmath2 below 90 min . once the @xmath1@xmath30 and @xmath1@xmath2 relations are found , it is straightforward to calculate the temporal change of the donor mass by solving @xmath166 .
we assume that donors are 1 gyr old and have a fixed initial mass of @xmath167 when they emerge from the period gap .
since cvs with @xmath168 h are predominantly dne @xcite , we further assume that the masses of wds are constant at @xmath43 throughout cv evolution .
figure [ pic_t_m ] plots as dashed and solid lines the evolution of donor mass and orbital period calculated from the semi - empirical @xmath2 for the b1 and b5 sequences , respectively , as functions of time @xmath169 elapsed since @xmath16 ( i.e. , @xmath170 gyr ) .
the dot - dashed lines show @xmath171 and @xmath172 from the mass - loss rate of @xcite for which @xmath173 and @xmath174 initially at @xmath16 .
the changes in the donor mass and orbital period in the numerical model of @xcite are in general slower because of a lower mass - loss rate .
our semi - empirical results indicate that @xmath30 decreases from @xmath167 to @xmath52 in @xmath48 gyr and that the minimum period is reached in @xmath48@xmath40 gyr after @xmath16 , without no essential dependence on the adopted stellar age .
on the other hand , the numerical model of @xcite predicts that the same decrement of @xmath30 occurs in @xmath175 gyr , while it takes the donors @xmath49 gyr to reach @xmath37 .
although uncertainties in the stellar ages make it difficult to follow @xmath171 and @xmath172 accurately below the mhbm , the close agreement between the results based on the b1 and b5 sequences suggests that the ages of cv secondaries with @xmath176 are less than 2 gyr .
mass transfer in cvs below the period gap is considered to be driven by gravitational wave radiation @xcite .
while equation ( [ eq_dm_def ] ) is useful to obtain @xmath2 from the empirical mrr independent of specifics of mass transfer , it does not provide any clue regarding what drives mass transfer in cvs . in this subsection , we make use of the stationary mass - transfer model of @xcite and check whether the semi - empirical @xmath2 calculated in [ ssec_dm ] is consistent with the angular momentum loss via emission of gravitational waves .
@xcite presented a simple analytical model in which the mass transfer from a donor is treated as a stationary isothermal , subsonic flow of gas through the inner lagrangian point .
taking allowance for the finite scale height of the donor photosphere and considering the temporal changes in the volume radii of the donor and the roche lobe , @xcite showed that the stationary mass - transfer rate is given by @xmath177,\ ] ] where @xmath178 is the angular momentum loss timescale , @xmath179 is the mass - radius exponent of the critical roche lobe .
note that we neglect the nuclear - timescale term in equation ( [ eq_dm_ritter ] ) since it is much longer than the other timescales involved . in the case of dne
for which the mass transfer is non - conservative , @xmath179 is given by @xmath180 @xcite . if the gravitational wave emission is the dominant mechanism for the shrinkage of the orbital radius , @xmath178 can be set equal to @xmath181 defined as @xmath182 @xcite .
we use the empirical mrr of @xcite and solve equation ( [ eq_dm_ritter ] ) with @xmath183 for @xmath184 .
the resulting @xmath1@xmath184 relations are plotted in figure [ pic_dm_p ] as dashed and solid lines in green color for the b1 and b5 sequences . respectively .
the adopted stellar ages make a small difference in @xmath184 , since the first term in the square brackets of equation ( [ eq_dm_ritter ] ) is insensitive to the isochrones when @xmath185 mhbm , while it becomes much smaller than the second term when @xmath186 mhbm .
figure [ pic_dm_p ] shows that @xmath184 agrees , within errorbars , with @xmath2 for both b1 and b5 sequences before reaching the minimum period .
this verifies that our semi - empirical mass - loss rate is consistent with the angular momentum loss due to emission of gravitational waves , at least for pre - bouncers .
beyond @xmath37 , @xmath2 for the b1 sequence appears to match @xmath184 better than the b5 sequence , suggesting that the cv ages are presumably less than 5 gyr .
since the mass - loss rate can not be observed directly , we are only able to compare our semi - empirical results with the accretion rates estimated by means of other indicators . for this purpose ,
we select 10 short - period , eclipsing cvs from @xcite , @xcite , and @xcite .
table [ ta_ecl ] represents a compilation of our sample cvs , with the parameters in columns ( 2 ) to ( 7 ) taken from the referred papers . to our knowledge , these are the only eclipsers with @xmath1 below 130 min , for which the reliable data for the binary parameters such as @xmath1 , @xmath42 , @xmath30 , and @xmath9 are all available . to infer the mass - accretion rates for the sample cvs , we use
@xmath187 where @xmath9 and @xmath42 are the effective temperature and mass of a wd and @xmath188 is the time - averaged accretion rate , taken from @xcite .
since the gravitational energy released by the accreted material onto the surface of a wd is deposited at shallow depths and soon radiated away , the quiescent surface luminosity should be proportional to the mass - accretion rate , leading to @xmath189 ( e.g. , @xcite ) .
filled circles with errorbars in figure [ pic_dm_te_ecl]_a _ plot @xmath188 for the sample cvs calculated from equation ( [ eq_town_2009 ] ) ; these are also listed in column ( 8) of table [ ta_ecl ] .
it can be seen that @xmath188 is broadly in line with our semi - empirical @xmath2 for both b1 and b5 sequence .
this suggests that our semi - empirical @xmath1@xmath2 relation is realistic , provided that @xmath190 in these cvs during accretion disk quiescence .
@xcite noted that @xmath188 significantly declines towards @xmath37 for @xmath191 min . since the accretion rate decreases with decreasing @xmath2 , this downward tendency of @xmath188 with @xmath1 is in agreement with the behavior of the semi - empirical @xmath2 .
donors in short - period cvs are very faint objects , making it difficult to measure their effective temperatures @xmath0 directly from observations , especially in the brown dwarfs regime : in our sample , only two systems ( ht cas and dv uma ) have measured @xmath0 available in the @xcite catalogue
. one way to estimate @xmath0 is to use the strong dependence of @xmath2 on @xmath0 . assuming that @xmath190 during the quiescence phase of dne , we combine equations ( [ eq_dm_def ] ) and ( [ eq_town_2009 ] ) to derive @xmath192 since donors with mass below the mhbm have @xmath85
, @xmath193 depends only on the orbital period and the effective mass - radius exponent . for masses above the mhbm , on the other hand , an accurate estimation of @xmath0 requires to consider the effect of @xmath194 term , which can not be determined observationally . using the method described in
[ ssec_lg ] , we calculate @xmath195 for the sample cvs listed in table [ ta_ecl ] . for this , we take @xmath196 for systems with @xmath197 and @xmath198 for @xmath199 in accordance with the adopted empirical mrr .
the bloating factor of each system is calculated relative to the 1-gyr isochrone .
we then calculate @xmath0 from equation ( [ eq_t1_t2 ] ) and plot the results in figure [ pic_dm_te_ecl]_b _ as filled circles with errorbars ; these are also tabulated in column ( 9 ) in table [ ta_ecl ] .
it is remarkable that the estimated effective temperatures , except for sdss 1507 and dv uma , are very close to the semi - empirical @xmath1@xmath0 relation , suggesting that equation ( [ eq_t1_t2 ] ) is in fact a good way to estimate @xmath0 when it can not be measured observationally .
a small orbital period of sdss 1507 implies that this system is most likely formed directly from a wd / brown dwarf pair @xcite or a member of the old halo @xcite , rather than a conventional dn . in the case of dv uma , @xmath200 k calculated from equation ( [ eq_t1_t2 ] ) is @xmath201 smaller than the observed effective temperature @xmath202 k , corresponding to a spectral type of m4.5@xmath203 , from the @xcite catalogue .
such a big difference in @xmath0 can not be due to ambiguity in @xmath204 or @xmath194 terms since both are close to unity for this object .
@xcite noted that the best - constrained quantity from a measured @xmath9 is the accretion rate per unit area , @xmath205 , while the subsequent estimation of @xmath188 from @xmath206 requires a certain mrr of wds , which is poorly known for cvs .
the wd mass in dv uma is highest among the sample , and it is highly probable that the mrr for massive wds differs from @xmath207 adopted by @xcite . based on the donor masses , @xcite claimed that sdss 1035 , sdss 1501 , sdss 1507 , and sdss 1433 possess post - period minimum cvs .
although period bouncers should have low @xmath30 , the precise mass of a particular secondary does not signify its evolutionary history or status @xcite .
the evolutionary status of such objects can be clarified more clearly by their positions , for instance , on the @xmath1@xmath0 or @xmath1@xmath2 diagram where the boomerang effect can be seen .
figure [ pic_dm_te_ecl ] suggests that sdss 1501 and sdss 1433 among the sample of @xcite have not reached @xmath37 yet .
this result does not depend on the stellar age , since these have @xmath208 for which the semi - empirical mass - loss rate is insensitive to the adopted isochrone .
while the evolutionary status of sdss 1035 is unclear , the fact that the @xmath1@xmath2 dependence for the b1 sequence reconciles well with that due to gravitational wave emission indicates that it is likely a post - bounce cv .
@xcite argued that the tidal and the rotational distortions do not change the value of @xmath34 appreciably because the internal structure of cv secondaries is only weakly affected by such perturbations . to show this , they compared the effective temperatures of the distorted and undistorted stars and found almost no change in @xmath0 .
this result is totally expected since short - period cvs are fully convective stars . as we showed in [ ssec_te ]
, the hayashi theory gives that the reaction of @xmath0 on the bloating of a donor is small for isolated , fully convective stars .
we also show in appendix [ ap_hayashi_mod ] that the influence of an isentropic expansion on @xmath0 caused by tidal and rotational distortions can be neglected in most practical cases .
all of these suggest that the effective temperature does not reliably measure the amount of the stellar bloating , and consequently can not be used as an indicator of the degree of tidal and rotational distortions .
the mass - loss rate of donors is one of the most important parameters that govern the evolution of short - period cvs . since the mass - loss rate is not directly observable , most previous work employed either purely numerical methods or semi - empirical approaches that use the observed properties of cv accretors rather than donors .
yet , the confirmation of the resulting mass - loss rates has to be made . in this work ,
we use the empirical mrr of short - period cvs constructed by @xcite as an input parameter and derive the semi - empirical relationship between the orbital period and mass - loss rate .
we assume that the bloating of cv donors is caused mainly by the mass loss and associated thermal relaxation processes .
we also calculate the responses of the effective temperature and luminosities on the bloating of the cv secondaries .
we discuss our semi - empirical mass - loss rate in comparison with those from the previous numerical and semi - empirical studies , and in terms of estimating the effective temperature of the donors .
our main results are summarized as follows .
the semi - empirical @xmath1@xmath2 dependence shows that the mass - loss rate is essentially non - stationary . for @xmath6 min ,
@xmath2 is in the range of @xmath209@xmath160 and decreases with decreasing @xmath1 relatively slowly as @xmath161 .
below 90 minutes , on the other hand , @xmath2 sharply declines towards @xmath37 as a result of a strong dependence of @xmath2 upon @xmath0 , emulating the @xmath1@xmath0 relation .
the @xmath1-dependence and amplitudes of the semi - empirical @xmath2 are significantly different from those predicted by the purely numerical models of @xcite .
the main reason for the discrepancies is small donor radii in the latter .
donors in short - period cvs are out of thermal equilibrium with @xmath164 .
thermal relaxation processes change the slope of the @xmath1@xmath2 and @xmath1@xmath0 diagrams significantly above @xmath37 , making the boomerang shape less pronounced . due to the sharp decline of @xmath0 and @xmath2 below 90 min ,
the observational detection of the boomerang shape on the @xmath1@xmath2 and @xmath1@xmath0 is highly unlikely .
the semi - empirical @xmath1@xmath2 relation on the basis of the 1-gyr isochrone is consistent with that estimated from the mass - transfer model of @xcite with the angular momentum carried dominantly by gravitational waves .
our semi - empirical @xmath1@xmath2 relation predicts that after emerging from the period gap , donors lose mass from @xmath167 to @xmath52 in @xmath48 gyr , and reach the minimum period in @xmath482 gyrs , suggesting that the ages of cv secondaries with @xmath210 are less than 2 gyrs .
\3 . for 10
selected eclipsing cvs ( listed in table [ ta_ecl ] ) for which the reliable observational data are available , the accretion rates inferred from the observed effective temperatures of wds are in good agreement with our semi - analytic mass - loss rates , indicating that our semi - empirical @xmath1@xmath2 relation is realistic .
this also suggests that the effective temperature of a cv secondary can be estimated from the effective temperature of its wd primary through equation ( [ eq_t1_t2 ] ) .
when analyzed on the @xmath1@xmath2 or @xmath1@xmath0 plane , two cvs ( sdss 1501 and sdss 1433 ) that were previously considered as post - bouncers by @xcite are most likely to be systems before reaching the minimum period .
the authors are grateful to the referee , dr . j. patterson , for a thoughtful report .
this work was supported by the national research foundation of korea ( nrf ) grant funded by the korean government ( mest ) , no .
2009 - 0063616 .
even without considering the effect of mass loss , @xcite showed that the tidal and rotational distortions alone make donors in cvs larger by about @xmath211@xmath212 for @xmath213 polytropes and @xmath214@xmath56 for @xmath215 polytropes , depending on the mass ratio . in this appendix
, we evaluate the enlargement factor for polytropes with arbitrary @xmath216 .
consider a polytropic star with index @xmath89 , mass @xmath30 , and pressure constant @xmath72 .
its radius is @xmath51 in isolation .
the mrr of the unperturbed polytropes is given by @xmath217 where @xmath218 is the dimensionless pressure constant that varies weakly with @xmath89 @xcite .
when it is situated in a cv , the presence of tidal and rotational perturbations causes it to expand or contract , depending on @xmath89 , by adjusting the internal structure to restore hydrostatic equilibrium .
let the new equilibrium be achieved at the volume radius @xmath25 with the corresponding dimensionless pressure constant @xmath219 . since the global internal adjustment occurs in the dynamical timescale , shorter than the thermal timescale and the mass - loss time scale , one may assume that the specific entropy of the gas inside the donor does not change much under the influence of the distortional perturbations . for isentropic responses ( i.e.
, @xmath72 is unchanged ) , the enlargement factor is given simply by @xmath220 we use a self - consistent field method outlined in @xcite to construct detailed internal structure of polytropes in hydrostatic equilibrium .
we then calculate @xmath218 and @xmath219 for polytropes in critical ( i.e. , roche - lobe filling ) configurations by varying @xmath89 and the mass ratio @xmath100 .
figure [ pic_npl ] plots the resulting relative changes , @xmath221 , in the stellar size as a function of @xmath89 for @xmath90 .
the behavior of @xmath88 with @xmath30 ( or with the mass ratio @xmath222 for @xmath43 ) for @xmath70 polytropes is plotted in figure [ pic_m_al ] as a dotted line .
figure [ pic_npl ] shows that @xmath223 for @xmath224 or @xmath225 , indicating that the perturbations make the donors bigger ; the increase in the equatorial radius is larger than the decrease in the polar radius . on the other hand , critical polytropes with @xmath226
have @xmath227 , since the perturbations decrease the polar radius much more than increasing the equatorial radius .
notice a discontinuity at @xmath228 for which one can not determine the stellar enlargement factor . for these polytropes ,
the central density is proportional to @xmath229 and the mass becomes independent of @xmath25 , making it impossible to constrain the mass from its radius ( e.g. , @xcite ) .
@xmath230 diverges to either positive or negative infinity as @xmath231 , indicating that a small change of @xmath89 near 3 leads to a big difference in @xmath230 .
this suggests that one should be cautious when treating @xmath228 polytropes numerically . using sph simulations
, @xcite found that the critical configurations have an enlargement factor @xmath232@xmath233 for @xmath228 polytropes when @xmath234@xmath235 .
this is presumably due to some ( unavoidable ) numerical errors in the sph simulations that somehow change their @xmath228 polytropes to @xmath236@xmath237 ones `` effectively '' , resulting in @xmath238@xmath239 .
in [ ssec_te ] , we describe how the effective temperature @xmath0 of a fully - convective donor varies due to a homologous expansion using the hayashi theory . here
we study how @xmath0 responds to an isentropic expansion .
we assume that the pressure @xmath240 in the whole stellar interiors below photosphere depends on temperature as @xmath241 where @xmath242 is the mean molecular weight and @xmath243 is the gas constant .
the photosphere is determined at the radius where the optical depth is @xmath244 .
if the radiative pressure is unimportant , the photospheric pressure , @xmath245 , is given by @xmath246 where @xmath247 is the opacity and @xmath248 is the mean surface gravity . using a simple opacity law of the form @xmath249
, we obtain @xmath250 since we assume that the donor is in hydrostatic equilibrium , the photospheric pressure should equal the interior pressure at the interface between the convective interiors and the radiative photospheric layer . assuming @xmath251 , equations ( [ eq_p_int_def ] ) and ( [ eq_p_ph_def ] ) are combined to yield @xmath252 in contrast to the hayashi theory , the effective temperature _ decreases _ due to the donor expansion .
nevertheless , @xmath253 for @xmath254 , @xmath255 , and @xmath256 , typical values for cv secondaries , so that the change in @xmath0 due to the tidal and rotational distortions are still very small and can thus be neglected in practice .
andronov , n. , pinsonneault , m. , sills , a. 2003 , , 582 , 358a baraffe , i. , chabrier , g. , allard , f. , & hauschildt , p. h. 1998 , , 337 , 403 baraffe , i. , chabrier , g. , allard , f. , & hauschildt , p. h. 2002 , , 382 , 563 baraffe , i. , chabrier , g. , barman , t. s. , allard , f. , & hauschildt , p. h. 2003 , , 402 , 701 barker , j. , & kolb , u. 2003 , , 340 , 623 burrows , a. , & liebert , j. 1993 , revmodphys , 65 , 301 chandrasekhar , s. 1933 , , 93 , 390 chandrasekhar , s. 1939 , an introduction to the study of stellar structure ( chicago univ . press ) dantona , f. , mazzitelli , i. , ritter , h. 1989 , , 225 , 391 eggleton , p. p. 1983
, , 268 , 368 hansen , c. j. , & kawaler , s. , d. 1995 , stellar interiors physical principles , structure , and evolution , ed .
appenzeller , i. , harvit , m. , kippenhahn , r. , strittmatter , p. , a. , ( springel - verlag , new - york , inc . ) howell , s. b. , nelson , l. a. , & rappaport , s. 2001 , , 550 , 897 king , a. r. 1988 , , 29 , 1 kippenhahn , r. , trimble , v. , & weigert , a. 1990 , stellar structure and evolution , ed .
harvit , m. , kippenhahn , r. , trimble , v. , zahn , j .-
p . , ( springel - verlag , berlin heidelberg ) knigge , c. 2006 , , 373 , 484 kolb , u. , & baraffe , i. 1999 , , 309 , 1034 kopal , z. 1972 , adv . astron .
. 9 , 1 landau , l. d. , & lifshitz , e. m. 1975 , the classical theory of fields , 4th edn .
( pergamon , new york ) littlefair , s. p. , dhillon , v. s. , marsh , t. r. , gnsicke , b. t. , baraffe , i. , & watson , c. a. 2008 , , 388 , 1582 littlefair , s. p. , dhillon , v. s. , marsh , t. r. , gnsicke , b. t. , baraffe , i. , & watson , c. a. 2007 , , 381 , 827 paczynski , b. 1981 , acta astron . , 31 , 1 patterson , j. , et al .
1984 , , 54 , 443 patterson , j. , et al .
2005 , , 117 , 1204 patterson , j. 2009 , , submitted ( arxiv:0903.1006 ) patterson , j. , thorstensen , j. r. , & knigge , c. 2008 , , 120 , 510 rappaport , s. , joss , p. c. , & webbink , r. f. 1982 , , 254 , 616 renvoiz , v. , baraffe , i. , kolb , u. , & ritter , h. 2002 , , 389 , 485 rezzolla , l. , uryu , k. , & yoshida , s. 2001 , , 888 ritter , h. 1988 , , 202,93 ritter , h. 1996 , in evolutionary processes in binary stars , proceedings of the nato advanced study institute , cambridge , ed .
wijers , r. , & davies , m. , 477 , 223 ritter , h. , & kolb , u. 2003 , , 404 , 301 sirtokin , f. v. , & kim , w .- t .
2009 , , 698 , 715(sk09 ) sion , e. , m. , 1995 , , 438 , 876 soberman , g. e. , phinney , e.s .
, & heuvel e.p.j .
1997 , , 327 , 620 stehle , r. , ritter , h. , & kolb , u. 1996 , , 279 , 581 townsley d. m. , & bildsten l. 2003 , , 596 , l227 townsley d. m. , & bildsten l. 2004 , , 600 , 390 townsley , d. m. , & gnsicke , b. t. 2009 , , 693 , 1007 gnsicke , b. t. et al .
2009 , , 397 , 2170 uryu , k. , & eriguchi , y. 1999 , , 303 , 329 urban , j. a , & sion e. m. 2006 , , 642 , 1029 cccccccccccccc 0.040&0.099 & 82.769 & 3.50&-4.625&1273&-11.10&0.24&&13.08&-5.513 & 763&-11.99&0.24 + 0.044&0.101 & 81.252 & 3.10&-4.435&1407&-10.94&0.23 & & 9.74&-5.317 & 846&-11.83&0.23 + 0.048&0.103 & 79.898 & 2.80&-4.293&1513&-10.83&0.23 & & 7.31&-5.167 & 914&-11.71&0.23 + 0.053&0.105 & 78.676 & 2.55&-4.194&1588&-10.76&0.23 & & 5.47&-5.011 & 992&-11.58&0.23 + 0.057&0.107 & 77.565 & 2.33&-4.121&1645&-10.73&0.24 & & 4.08&-4.880&1062&-11.49&0.24 + 0.063&0.108 & 76.021 & 2.05&-3.945&1804&-10.67&0.29 & & 2.64&-4.507&1305&-11.23&0.29 + 0.065&0.111 & 77.472 & 1.95&-3.828&1905&-10.61&0.32 & & 2.30&-4.316&1439&-11.09&0.32 + 0.069&0.116 & 79.730 & 1.80&-3.679&2035&-10.56&0.40 & & 1.87&-4.108&1590&-10.99&0.40 + 0.075&0.121 & 82.611 & 1.70&-3.481&2226&-10.39&0.41 & & 1.72&-3.649&2021&-10.55&0.41 + 0.078&0.124 & 84.036 & 1.65&-3.384&2327&-10.33&0.44 & & 1.67&-3.461&2226&-10.39&0.43 + 0.082&0.128 & 86.094 & 1.59&-3.255&2466&-10.21&0.44 & & 1.60&-3.282&2427&-10.23&0.44 + 0.086&0.133 & 88.096 & 1.56&-3.146&2584&-10.11&0.45 & & 1.57&-3.156&2570&-10.12&0.44 + 0.090&0.137 & 90.046 & 1.53&-3.063&2671&-10.04&0.46 & & 1.54&-3.064&2668&-10.04&0.45 + 0.095&0.141 & 91.948 & 1.50&-2.987&2751 & -9.98&0.46 & & 1.52&-2.986&2751 & -9.98&0.46 + 0.099&0.144 & 93.805 & 1.47&-2.925&2812 & -9.94&0.47 & & 1.49&-2.924&2813 & -9.93&0.47 + 0.103&0.148 & 95.619 & 1.44&-2.870&2865 & -9.90&0.48 & & 1.46&-2.869&2866 & -9.89&0.48 + 0.107&0.152 & 97.394 & 1.41&-2.821&2908 & -9.86&0.49 & & 1.43&-2.820&2910 & -9.86&0.49 + 0.111&0.156 & 99.132 & 1.38&-2.778&2946 & -9.84&0.51 & & 1.40&-2.777&2948 & -9.83&0.50 + 0.116&0.160&100.835 & 1.36&-2.736&2983 & -9.81&0.52 & & 1.38&-2.735&2984 & -9.81&0.52 + 0.120&0.163&102.504 & 1.34&-2.699&3012 & -9.79&0.53 & & 1.36&-2.698&3014 & -9.79&0.53 + 0.124&0.167&104.142 & 1.32&-2.666&3037 & -9.77&0.54 & & 1.34&-2.665&3038 & -9.77&0.54 + 0.128&0.170&105.751 & 1.30&-2.636&3057 & -9.76&0.56 & & 1.32&-2.636&3058 & -9.76&0.56 + 0.132&0.174&107.331 & 1.28&-2.605&3080 & -9.74&0.57 & & 1.30&-2.605&3081 & -9.74&0.57 + 0.137&0.177&108.885 & 1.26&-2.575&3104 & -9.73&0.58 & & 1.28&-2.574&3105 & -9.72&0.58 + 0.141&0.181&110.413 & 1.24&-2.547&3124 & -9.71&0.59 & & 1.26&-2.546&3125 & -9.71&0.59 + 0.145&0.184&111.916 & 1.22&-2.521&3141 & -9.70&0.61 & & 1.24&-2.520&3143 & -9.70&0.60 + 0.149&0.188&113.396 & 1.20&-2.497&3155 & -9.69&0.62 & & 1.22&-2.496&3157 & -9.69&0.62 + 0.154&0.191&114.854 & 1.18&-2.472&3174 & -9.68&0.64 & & 1.20&-2.471&3176 & -9.68&0.63 + 0.158&0.194&116.291 & 1.16&-2.447&3191 & -9.67&0.65 & & 1.18&-2.446&3193 & -9.67&0.65 + 0.162&0.198&117.706 & 1.15&-2.425&3206 & -9.66&0.67 & & 1.16&-2.424&3208 & -9.66&0.66 + 0.166&0.201&119.103 & 1.13&-2.403&3219 & -9.66&0.68 & & 1.15&-2.403&3220 & -9.65&0.68 + 0.170&0.204&120.480 & 1.11&-2.383&3230 & -9.65&0.70 & & 1.13&-2.383&3231 & -9.65&0.70 + 0.175&0.207&121.839 & 1.10&-2.365&3240 & -9.65&0.72 & & 1.11&-2.364&3241 & -9.65&0.72 + 0.179&0.211&123.180 & 1.08&-2.345&3252 & -9.64&0.73 & & 1.09&-2.344&3253 & -9.64&0.74 + 0.183&0.214&124.505 & 1.06&-2.326&3263 & -9.63&0.74 & & 1.07&-2.325&3264 & -9.64&0.76 + 0.187&0.217&125.813 & 1.05&-2.308&3272 & -9.62&0.75 & & 1.05&-2.307&3274 & -9.64&0.78 + 0.191&0.220&127.105 & 1.03&-2.291&3281 & -9.61&0.75 & & 1.03&-2.290&3283 & -9.64&0.81 + 0.196&0.223&128.382 & 1.01&-2.275&3289 & -9.60&0.76 & & 1.02&-2.274&3290 & -9.64&0.84 + 0.200&0.226&129.563 & 1.00&-2.260&3295 & -9.60&0.78 & & 1.00&-2.260&3296 & -9.65&0.88 + [ ta_seq ] lrcccccccc sdss 1507 & 66.6119&@xmath257&@xmath258&@xmath259&@xmath260&@xmath261&@xmath262&@xmath263&3 + + sdss 1433 & 78.1066&@xmath264&@xmath265&@xmath266&@xmath267&@xmath268&@xmath269&@xmath270&3 + + sdss 1501 & 81.8513&@xmath271&@xmath272&@xmath273&@xmath274&@xmath275&@xmath276&@xmath277&3 + + sdss 1035 & 82.0896&@xmath278&@xmath279&@xmath280&@xmath281&@xmath282&@xmath283&@xmath284&3 + + sdss 1502 & 84.8298&@xmath285&@xmath286&@xmath287&@xmath288&@xmath289&@xmath290&@xmath291&3 + + sdss
0903 & 85.0659&@xmath292&@xmath293&@xmath294&@xmath295&@xmath296&@xmath297&@xmath298&3 + + sdss 1227 & 90.6610&@xmath299&@xmath300&@xmath301&@xmath302&@xmath303&@xmath304&@xmath305&3 + + oy car & 90.8928&@xmath306&@xmath307&@xmath308&@xmath309&@xmath310&@xmath311&@xmath312&1,2,3 + + ht cas & 106.0560&@xmath313&@xmath314&@xmath315&@xmath316&@xmath317&@xmath318&@xmath319&1,2 + + dv uma & 128.2800&@xmath320&@xmath321&@xmath322&@xmath323&@xmath324&@xmath325&@xmath326&1,2 + + [ ta_ecl ] | the mass - loss rate of donor stars in cataclysmic variables ( cvs ) is of paramount importance in the evolution of short - period cvs .
observed donors are oversized in comparison with those of isolated single stars of the same mass , which is thought to be a consequence of the mass loss . using the empirical mass - radius relation of cvs and the homologous approximation for changes in effective temperature @xmath0 , orbital period @xmath1 , and luminosity of the donor with the stellar radius
, we find the semi - empirical mass - loss rate @xmath2 of cvs as a function of @xmath1 .
the derived @xmath2 is at @xmath3@xmath4 @xmath5 and depends weakly on @xmath1 when @xmath6 min , while it declines very rapidly towards the minimum period when @xmath7 min , emulating the @xmath1@xmath0 relation . due to strong deviation from thermal equilibrium caused by the mass loss ,
the semi - empirical @xmath2 is significantly different from , and has a less - pronounced turnaround behavior with @xmath1 than suggested by previous numerical models .
the semi - empirical @xmath1@xmath2 relation is consistent with the angular momentum loss due to gravitational wave emission , and strongly suggests that cv secondaries with @xmath8 are less than 2 gyrs old .
when applied to selected eclipsing cvs , our semi - empirical mass - loss rates are in good agreement with the accretion rates derived from the effective temperatures @xmath9 of white dwarfs , suggesting that @xmath2 can be used to reliably infer @xmath0 from @xmath9 . based on the semi - empirical @xmath2 ,
sdss 1501 and 1433 systems that were previously identified as post - bounce cvs have yet to reach the minimal period . |
laser ranging to earth artificial satellites ( slr ) was initiated in 1964 after launch of the first geodetic / geodynamical satellite beacon - b . since that time satellite laser ranging ( slr ) technique have being widely used for geodynamical and geophysical researches .
the primary fields of investigations used slr observations are earth rotation , maintenance of the terrestrial reference frame , tectonic motion , earth crust deformations , geopotential with its spatial and temporal variations , tides , movement of geocenter , support of satellite geophysical missions ( such as satellite altimetry ) , global time transfer , and others .
detailed analysis of slr contribution to earth sciences can be found in ( tapley _ et al .
_ 1993 ) . after the launching of the lageos satellite in may 1976 , slr became one of the main techniques for investigations of the earth rotation , and during over twenty years slr technique
have being used for determination of erp .
most of results contributed to iers are based on analysis of observations of lageos 1&2 satellites collected at the global tracking network of about 40 stations .
slr provides high precision series of xp , yp , and lod .
some analysis centers compute also ut that allow to densify universal time series in combination with vlbi data .
importance of ilrs as one of the main method to study the earth led , naturally , to the establishment of the international laser ranging service ( ilrs ) in 1998 that coordinates now scientific activity slr ( and llr ) , chiefly in the framework of iag and iers projects .
in this paper we will focus only on slr contribution to investigation of polar motion in accordance with topics of the conference .
slr observations of lageos and erp derived from these observations are available from may 1976 .
however , only in 198081 after significant improvement of range technique slr erp series achieved accuracy required for investigation of polar motion ( pm ) .
figure [ fig : ac_contr ] summarizes information from iers annual reports for 19781997 concerning use of submitted slr series in iers yearly solutions .
one can see that beginning from 1983 slr erp series is regularly used for iers combination .
it s also seen that the center for space research , university of texas at austin ( csr ) provides most long - time spanned and stable pm series .
= at present 5 analysis centers submit operational solutions with 2 - 15 days delay , and about 10 analysis centers yearly contribute final ( up to 23 years ) erp series .
most long series available for analysis are listed in table [ tab : long_series ] .
during long history of using slr technique to study pm drastic improvement in range precision was achieved .
it s interesting to see how this technology development affected accuracy of pm series obtained from slr observations .
figure [ fig : csriaa ] shows differences between slr series obtained at the csr and the institute of applied astronomy , st.petersburg ( iaa ) and iers combined solution eop(iers)c04 .
one can clearly see in figure [ fig : csriaa ] that replacement of the first generation ranging equipment by the second generation one about 1980 and its further replacement by the third generation units leaded to significant improvement of accuracy of pm solution .
however , in spite of improvement of ranging technique and implementation the fourth generation equipment , precision of pm series derived from slr observations remains the same during last 1012 years .
we will try to discuss the problem below . in 1998 iers
central bureau had derived final iers combination in two step : on the first stage combined vlbi , slr , and gps series was computed and then these was used for final iers combination .
table [ tab : acc_c04 ] copied from the 1997 iers annual report shows that precision of slr and gps combinations is practically the same .
unfortunately , this is not so for operational solutions .
table [ tab : acc_usno ] compiled on basis of ` gpspol.asc ` files produced by usno along with bulletin a issues contains statistics of series used for iers rapid service .
of course , errors in pm obtained there depend on weighting applied to combined series , but it already clear that accuracy of individual slr series is worse than vlbi and gps ones , and delay of slr contribution is much worse than gps one .
nevertheless , one can see that at least two centers can produce operational solution with delay about 2 days .
if all slr analysis centers would provide such a solution on regular basis , it will allow to have , in principle , combined slr solution with accuracy 0.10.15 mas and delay 1.52 days .
during long time slr was the one of the main methods of determination of pm and densification of pm series based on vlbi observations .
however , lately slr erp series are inferior to gps ones in quality ( accuracy , density , delay of operational solution ) of results
. it would be very important to understand existing problems in slr observations and analysis procedures and discuss possible ways to solve them .
the first problem with slr observations is that slr technique is an one - object one .
this means that , unlike gps , slr station can observe only one object in time .
hence , planning of observations and priority politics plays substantial role in acquisition of observations of satellites on which pm determination is based , especially keeping in mind that slr observations are rather expensive and number of ranging units are limited . at present four operational satellites
seem most suitable for investigation of earth rotation ( as well as tectonic motion and long - term temporal variations in various geophysical parameters ) two lageos satellites and two etalon satellites .
both lageos and etalon satellite was launched to long - time stable orbits and have a low area to mass ratio .
their description is presented in figure [ fig : satellit ] .
figure [ fig : observ ] shows number of observations of these satellites during last 12 years and ilrs priorities .
one can see that number of observations of lageos satellites remains approximately the same during these years and number of observations of etalon satellite is too small to contribute seriously to analysis .
= 0.75 let us see in more details distribution of slr observations in stations and time .
figure [ fig : sta_map ] shows distribution of observations in stations for the period from sep 1983 till aug 1999 .
table [ tab : obs_sta ] contains list of stations contributed more than 2% of total number of observations for whole period and during the last year ( in parenthesis ) .
= 0.8 .distribution of observations in stations .
[ cols="^,<,<,^,<,<",options="header " , ] [ tab : daily_int ] it should be mentioned also that producing daily solution ( independently which method i d used ) provide operational solution with steady delay about 2 days , which solves the second problem mentioned above .
other serious problem is determination of ut from slr ( and other satellite observations ) .
it is well known that ut1 can not be separated from longitude of node of satellite orbit during parameter solution .
three methods are being used to solve this problem : * fixing longitude of node during parameter solution ( usually , during last iteration ) .
* analysis of node longitude series , forecasting it and use predicted values for operational solution . * integrating lod series to obtain independent free - running ut series with its possible correction for high - frequency variations from comparison with vlbi series . evidently , only the latter method can provide ( in principle ) independent result .
since that is not a subject of this paper , we will not stay on detailed analysis of this problem .
however , it is worth to mention here that significant improvement of slr ut series is also impossible without increasing of number of satellites involved in determination of erp . at last ,
use the same terrestrial reference frame for all slr solutions seems evident to achieve uniform solutions for combination .
use of itrf as terrestrial reference frame for by all analysis centers for their slr solutions provides more homogeneous series for slr combined solution .
realization of this or alternative analysis strategy could provide more uniform , accurate and operative slr erp series . after that , combining of all submitted series to final ilrs slr product seems reasonable and useful for further use for iers and other purposes .
we have not mention here such a serious problem as dependence of slr erp results on a priori values .
this is worth to perform special investigation for each method used for computation of erp at various analysis centers .
satellite laser ranging technique made and make a very valuable contribution to earth dynamics . in particular ,
very valuable contribution was made in investigation of pm . during many years
slr was one of the main methods of determination of polar motion and main method of densification of erp series obtained with vlbi . at present 5 analysis centers
submit operational solutions with 2 - 15 days delay , and about 10 analysis centers yearly contribute final ( up to 23 years ) erp series .
however , due to principal peculiarity of this method ( relatively expensive experiment , limited number of units , lack of capability of multi - satellite ranging , etc . )
quality of slr erp data remains the same during the last decade in spite of ranging precision improved by a factor of a thousand from a few meters to a few millimeters since the first slr experiments ( and by factor of about 10 during last 10 - 12 years ) .
this leads to decreasing of weight of slr solutions in the combined iers eop series .
it is evident that only substantial increasing of number of observations and satellites involved in investigation of earth rotation and improvement of distribution of observations in stations and time can help in improvement of the slr erp accuracy .
however , capacity of tracking network is practically exhausted .
on the other hand , in spite of gps provides determination of erp with impressive accuracy and delay , slr results are very important for combined iers solution for improvement of systematic accuracy of the final iers product .
analysis of precision of individual slr series shows that its combination can provide combined slr much more accurate and rapid series .
to achieve highest accuracy of combined slr erp product is necessary to solve problems discussed in section [ sect : analysis ] .
realization of this opportunities by ilrs would be very important for investigation of earth rotation because allow to save independent method of determination of pm and velocity of the earth rotation .
author is very grateful to scientific organizing committee for invitation to the meeting and financial support of this trip . | slr technique has being used for determination of erp during over twenty years .
most of results contributed to iers are based on analysis of observations of lageos 1&2 satellites collected at the global tracking network of about 40 stations .
now 5 analysis centers submit operative ( with 2 - 15 days delay ) solutions and about 10 analysis centers yearly contribute final ( up to 23 years ) erp series .
some statistics related to slr observations and analysis is presented and analyzed .
possible problems in slr observations and analysis and ways of its solution are discussed . |
understanding charge transport in graphene nanoribbons ( gnrs ) has attracted significant interest in recent years @xcite , in particular due to their potential application as integrated circuits @xcite and field - effect transistors @xcite , as bio - sensing devices @xcite , and for dna sequencing @xcite .
these unprecedented opportunities have been made possible by experimental advances in creating sub-10 nm wide gnrs @xcite , in engineering gnrs with specific electronic structures @xcite , in fabricating high purity samples @xcite , and in designing artificial molecular graphene @xcite .
moreover , due to their long mean - free path , gnrs are also an ideal system to explore the fundamental properties of out - of - equilibrium charge transport in nanoscopic quantum systems .
previous theoretical studies investigating the transport properties of gnrs have predominantly focused on the bias dependent conductance @xcite .
spatial current patterns were investigated in the vicinity of the fermi energy of half - filled gnrs in the wide - lead limit both for zero - magnetic field @xcite and for non - zero magnetic fields and disorder @xcite .
recently , however , it was argued @xcite that in nanoscopic networks with narrow constrictions and/or narrow leads , spatial current patterns emerge which are qualitatively different from those in the wide - lead limit .
in particular , these spatial current patterns exhibit clear signatures of quantum behavior : they ( a ) possess coherent
current riverbeds " , i.e. , spatial regions of large current density , whose widths are of the order of the fermi wave - length , and ( b ) are strongly dependent on boundary conditions , such as the position of the leads , the geometry and size of the network , as well as the gate voltage
. these results clearly suggest that understanding the quantum nature of local charge transport in gnrs , as reflected in the form of spatial current patterns , and its dependence on a gnr s geometry and size , or the presence of dephasing , and developing a method to visualize it , is of utmost importance for the further development of graphene based nanoelectronics and dna sequencing . in this article
, we address this open question by demonstrating how the spatial current patterns in gnrs are determined by the interplay between the gnrs geometry , aspect ratio and size , by the location and number of leads , and the presence of dephasing . by using the non - equilibrium keldysh green s function formalism @xcite
we find that the gnr s unconventional electronic structure @xcite is reflected not only in the bias dependence of the conductance , but also in the form of the spatial current paths .
in particular , we find that each of a gnr s discrete transmission resonances possesses a characteristic spatial current pattern ( hence we refer to these resonances also as _ current eigenmodes _ in analogy to the equilibrium eigenmodes in the local density of states of nanoscopic systems ) .
we demonstrate that the spatial form of these current patterns is determined by the number of degenerate states participating in the charge transport , which in turn allows us to predict how these current patterns evolve with increasing size or aspect ratio of a gnr .
we show that current patterns can include closed loops of circulating currents , as well as exhibit _ backflow _ , i.e. , the flow of charge through certain links opposite to the direction of the net charge flow @xcite .
we demonstrate that the spatial current paths are qualitatively different between leads attached to the zig - zag or armchair edges of the gnr and explore the form of charge transport parallel and perpendicular to edge states .
moreover , we show that while current patterns in certain 4-lead configurations can arise from the superposition of two 2-lead configurations , the conductance does not necessarily obey the superposition principle .
furthermore , we demonstrate that the quantum behavior of local charge transport can be visualized using scanning tunneling microscopy ( stm ) @xcite : it represents an essentially non - intrusive method to image spatial current paths in gnr with atomic resolution , providing important insight into the form of local charge transport .
finally , we investigate more realistic models including the effects of dephasing and next - nearest - neighbor hopping , and show that with decreasing dephasing time , the current patterns evolve smoothly from those of a ballistic quantum network to those of a classical resistor network .
these results provide important insight into the fundamental aspects of charge transport in nanoscopic graphene lattices .
to study electron transport in a graphene nanostructure , we consider two one - dimensional leads coupled to a finite @xmath0 honeycomb lattice with @xmath1 hexagonal cells in the armchair direction , and @xmath2 cells in the zig - zag direction , shown schematically in figure [ fig : ngl ] .
graphene nanoribbon ( gnr ) .
the leads are connected to gnr sites @xmath3 and @xmath4.,width=377 ] the graphene nanoribbon is described by the hamiltonian @xcite @xmath5 where @xmath6 is the electronic hopping matrix element between sites @xmath7 in the gnr , and @xmath8 creates an electron with spin @xmath9 at site @xmath10 . in what follows ,
we take only the nearest neighbor hopping element , @xmath11 , to be non - zero ; the effects of a next - nearest neighbor hopping @xmath12 will be discussed in sec .
[ sec : nnn ] . in order to account for the effects of dephasing
, we assume that the electrons interact locally ( with coupling constant @xmath13 ) with a phonon mode of energy @xmath14 , as described by the last two terms in equation ( [ eq : hamiltonian ] ) , where @xmath15 creates a phonon at site @xmath10 .
the coupling between the gnr and the leads is described by the hamiltonian @xmath16 where the primed sum runs over all sites @xmath10 and @xmath17 in the gnr and leads , respectively , that are coupled by a hopping element @xmath18 , and @xmath19 creates an electron with spin @xmath9 at site * l * in the leads .
below , we assume that each of the two leads is coupled to a single gnr site only , labeled @xmath3 and @xmath4 ( see figure [ fig : ngl ] ) . therefore , the only relevant property of the leads entering our calculations is the local green s function at the lead sites that are coupled to the gnr .
finally , for the purpose of imaging spatial current patterns in the gnr @xcite , we consider the tunneling of electrons from an stm tip to a single site * t * in the gnr , a process which is described by the hamiltonian @xmath20 where @xmath21 annihilates an electron with spin @xmath9 in the stm tip .
the spatial current patterns in a gnr are obtained by computing the current , @xmath22 , between adjacent sites @xmath10,@xmath23 in the gnr .
this current is induced by different chemical potentials , @xmath24 in the left and right leads , and given by @xcite @xmath25 \ , \label{eq : current}\ ] ] where @xmath26 is the full keldysh green s function between sites @xmath10 and @xmath23 which accounts for the electronic hopping within the gnr and between the gnr and the leads , as well as the electron - phonon interaction .
we next introduce a matrix notation such that the keldysh green s function matrix is given by @xmath27 , and its @xmath28 element , @xmath29 , is the keldysh green s function between sites in the gnr denoted by @xmath30 and @xmath31 . in the absence of an electron - phonon interaction ,
one finds @xmath32 with @xmath33 { \rm i m } \left [ { \hat g}^r(\omega ) \right]$ ] . here
, @xmath34 are the diagonal retarded , advanced and keldysh green s function matrices , respectively , containing the green s functions of the decoupled ( @xmath35 ) gnr sites and leads : for the gnr sites , we have @xmath36 with @xmath37 while for the leads , we assume the wide - band limit and take @xmath38 , yielding a constant density of states @xmath39 in the leads . finally , @xmath40 is a diagonal matrix containing the fermi - distribution functions and @xmath41 is the symmetric hopping matrix .
moreover , the local density of states at site @xmath10 in the gnr is obtained via @xmath42 . for the results shown below ,
we have used for numerical convenience @xmath43 , @xmath44 and @xmath45 , unless otherwise specified .
all spatial currents , @xmath46 , presented below are normalized to the largest current in the gnr . to account for the effects of the electron - phonon interaction for general @xmath47 and @xmath48 is computationally demanding and beyond the scope of this article .
however , since we are mainly interested in the effects of dephasing on the spatial current patterns , we can consider the high - temperature approximation introduced in ref .
@xcite . in this limit with @xmath49 , implying a large thermal population of the phonon mode , the calculation of the electronic self - energy correction is greatly simplified and computationally possible even for larger gnr sizes . retaining in the dyson equation only those terms that contain a factor of @xmath50 , the electronic self - energy in the self - consistent born approximation
is given by @xmath51 where @xmath30 denotes a site in the gnr and @xmath52 .
moreover , @xmath53 \left [ \delta(\omega + \omega_{0 } ) - \delta(\omega - \omega_{0})\right]\ ] ] is the keldysh phonon green s function , which we assume to remain unchanged by the electron - phonon interaction , and @xmath54 is the phonon bose distribution function ( we assume that the phonons remain in thermal equilibrium ) .
note that due to the coupling to local phonon modes , the electronic self - energy is entirely local .
a further simplification is achieved by considering the limit @xmath55 such that the self - energy , to leading order in @xmath56 , is given by @xmath57 as shown before @xcite , the solution of the dyson equation for the retarded green s function is then given by @xmath58^{-1 } \hat{g}^{r } \label{eq : se2}\ ] ] where @xmath59 is a superoperator introduced in ref .
@xcite which , when operating on a green s function matrix , returns the same matrix with all elements set to zero except for the diagonal elements in the matrix that correspond to sites in the gnr , e.g. , @xmath60_{ij } = \cases{g_{ij}^{\alpha}\delta_{ij } & if $ i$ is a site in the gnr\\ 0 & otherwise \ . \\}\ ] ] after self - consistently solving equation ( [ eq : se2 ] ) for @xmath61 , @xmath62 can be obtained in a closed expression via @xmath63 where the diagonal matrix @xmath64 is defined via @xmath65 _ { l } .\ ] ] here , the vector @xmath66 has components @xmath67 with @xmath68 being a diagonal matrix whose only non - zero elements @xmath69 are those where @xmath30 denotes one of the two leads .
the matrix @xmath70 contains the elements @xmath71
gnr , and ( b ) - ( h ) spatial current patterns for different @xmath73 . all spatial currents , @xmath46 , are normalized to the largest current in the gnr .
yellow arrow in
( f ) indicates a circulating current loop .
the leads are connected to gnr sites @xmath3 and @xmath4 . ]
we begin by discussing charge transport , and in particular the relation between the conductance and spatial current patterns , in an @xmath74 gnr in the absence of the electron - phonon interaction . in figure
[ fig : ngl11x5](a ) , we present the gnr s total conductance @xmath75 in the limit @xmath76 which is given by @xmath77 where @xmath78 is the bias midpoint with @xmath79 , @xmath80 is the leads local density of states , and @xmath81 is the non - local green s function between sites @xmath3 and @xmath4 where the current enters and exits the gnr ( see figure [ fig : ngl ] ) , respectively @xcite . due to its finite size , the gnr possesses discrete energy levels @xcite which for @xmath82 are given by @xmath83 ^ 2 = 1
\pm 4\cos\frac{\pi j}{2 ( n_{z } + 1 ) } \cos \frac{\kappa_{j}^{\pm}}{2 } + 4 \cos^{2}\frac{\pi j}{2(n_{z}+1)}\ ] ] where the @xmath84 are solutions of @xmath85 as a result , the conductance exhibits discrete transmission resonances whenever @xmath86 equals the energy a state whose wavefunction does not vanish at the site that the leads are coupled to . for each of these resonances ,
the current flowing through the gnr exhibits a different spatial pattern ( the resonances are therefore also referred to as current eigenmodes @xcite , in analogy to eigenmodes in the density of states ) .
the spatial current patterns [ as obtained from equation ( [ eq : current ] ) ] for a number of current eigenmodes are shown in figures [ fig : ngl11x5](b ) - ( h ) [ due to the particle - hole symmetry of the gnr s electronic structure , the conductance is symmetric around @xmath87 and the eigenmodes at @xmath88 exhibit identical current patterns ; we therefore restrict our discussion below to the case @xmath89 .
note that these eigenmodes can in general be accessed experimentally via gating of the gnr .
similar to prior results @xcite , we find that there exists current eigemodes which exhibit circulating current loops [ as indicated by a yellow arrow in figure [ fig : ngl11x5](f ) ] or which possess a net current through the gnr which is significantly smaller than some of the currents within the gnr [ see figure [ fig : ngl11x5](b ) ]
. of particular interest is the well - ordered current pattern that occurs for @xmath90 shown in figure [ fig : ngl11x5](d ) . to understand the spatial form of this current pattern ,
we note that charge transport for @xmath90 with leads attached to the zig - zag edge involves a @xmath91fold degenerate @xmath92 state ( not including @xmath93 states with vanishing amplitude along the zig - zag edge @xcite ) , whose wave - vectors are shown in figure [ fig : x5](a ) ( solid and open circles ) ; for comparison , we also present the @xmath92 equal energy contour for an infinitely large graphene sheet ( solid line ) .
only three of these 5 states [ green circles in figure [ fig : x5](a ) ] possess a non - zero wavefunction at the gnr sites @xmath3 and @xmath4 that the leads couple to in figure [ fig : ngl11x5 ] and thus contribute to the charge transport at @xmath90 .
the fermi velocity at @xmath92 is given by @xmath94 [ see figure [ fig : x5](a ) ] and thus parallel to the primitive lattice vectors of one of the triangular sublattices of graphene . as follows from figure
[ fig : ngl11x5](d ) , the current propagates generally along the same direction .
thus , the spatial current pattern is reminiscent of the motion of a ballistic particle that is specularly reflected off the walls of the gnr .
we find that this relation between the spatial current pattern and @xmath95 holds whenever there are degenerate states with the same @xmath95 participating in the charge transport .
the relation between the spatial current pattern for @xmath96 , the aspect ratio of the gnr , and the degeneracy of the @xmath92 state will be discussed in more detail in section [ sec : aspect ] .
it is noteworthy that , in the vicinity of @xmath96 , the spatial form of the current patterns evolves rapidly with increasing @xmath73 [ see figures [ fig : ngl11x5](c ) - ( e ) ] while for larger values of @xmath73 , the spatial form evolves more slowly [ see figures [ fig : ngl11x5](f ) - ( h ) ] .
finally , we note that for finite - sized gnrs , there exist two states near zero energy at @xmath97 , that are delocalized along the zig - zag edge , but localized in the armchair direction @xcite .
for @xmath98 , the gnr therefore exhibits strongly anisotropic transport properties in the armchair and zig - zag directions , as discussed in more detail in section [ sec : edge ] .
states ( closed and open circles ) in an @xmath74 gnr together with the @xmath92 equal energy contour in the brillouin zone of an infinite graphene sheet .
the wave - vectors of the three states possessing a non - zero wavefunction at site @xmath3 and @xmath4 are shown in green .
( b ) - ( f ) evolution of the spatial current patterns for @xmath96 in @xmath99 gnrs with increasing @xmath100 .
( g ) contour plot of the normalized density of states of the @xmath74 gnr at @xmath92.,width=604 ] in the ballistic limit , spatial current patterns in nanoscale systems sensitively depend on boundary conditions such as the location of the leads or the geometry or aspect ratio of the system @xcite . to investigate this issue in gnrs , we next study the dependence of spatial current patterns on the aspect ratio , @xmath101 of the gnr , as well as the location of the leads . to exemplify this dependence , we consider the current eigenmode at @xmath96 ,
since it does not only exhibit a spatially highly ordered structure , but also occurs in all gnrs considered below . in figure
[ fig : x5 ] , we present the evolution of the spatial current pattern with increasing @xmath100 in @xmath99 gnrs for @xmath96 .
we note for the discussion below that in all of these gnrs , there exist @xmath100 states at @xmath92 whose wavefunction vanishes at the sites that the leads are attached to @xcite ; those states are therefore irrelevant for the purpose of the charge transport considered here and will not be discussed below .
the spatial current patterns in the gnrs exhibit two types of characteristic forms .
in particular , in gnrs that satisfy @xmath102 with integer @xmath103 , the current follows the direction of the fermi velocity , @xmath95 , being specularly reflected off the walls of the gnr , as shown in figures [ fig : x5](b ) and ( e ) . in these gnrs
, there are @xmath104 states at @xmath92 whose wavefunctions do not vanish at sites @xmath3 and @xmath4 ; their wave - vectors are shown in figure [ fig : x5](a ) ( green circles ) , together with the @xmath92 equal - energy contour of an infinite graphene sheet .
with increasing @xmath103 , certain elements of the spatial current pattern are repeated [ for example , the current pattern in figure [ fig : x5](e ) with @xmath105 consists of two copies of the current pattern of figure [ fig : x5](b ) plus an additional current loop connecting these two elements ] .
in contrast , the gnrs shown in figure [ fig : x5](c),(d ) , and ( f ) do not satisfy the above requirement , possessing only a single @xmath92 state [ whose wave - vector is shown as the open circle in figure [ fig : x5](a ) ] .
consequently , these structures exhibit current patterns that are significantly more complex and possess a series of traits characteristic of quantum mechanical charge transport in nanostructures @xcite .
in particular , these current patterns exhibit circulating current loops that are isolated [ see red arrow in figure [ fig : x5](c ) ] and thus do not contribute to the net current through the gnr .
such current loops give rise to magnetic fields that could potentially be detected experimentally : we estimate that the magnetic field at the center of the current loop indicated by an arrow in figure [ fig : x5](c ) is approximately @xmath106 t .
in addition , these current patterns exhibit backflow : while a net current flows from the left to the right ( @xmath107 ) , there exist certain bonds in the gnr where the current flows from right to left ( and thus opposite to the applied bias ) , such as the ones indicated by a yellow arrow in figure [ fig : x5](d ) .
we find that circulating current loops and backflow occurs in all gnrs in which the current pattern does not simply reflect the direction of the fermi velocity , @xmath95 , as is the case in figures [ fig : x5](b ) and ( e ) .
finally , in figure [ fig : x5](g ) we present the local density of states for the @xmath74 gnr ( with leads attached ) at @xmath92 , which has no resemblance to the form of the spatial current pattern shown in figure [ fig : x5](e ) , a conclusion that was previously also drawn for nanoscale systems with a square lattice geometry @xcite .
this result stands in contrast to earlier findings relating either the spatial current pattern at @xmath73 directly to that of the modulus of the wave - function , @xmath108 @xcite ( which possesses the same spatial structure as the density of states , @xmath109 , since @xmath110 ) , or the spatial flow of electrons to the spatial form of @xmath111 @xcite .
more generally , our results demonstrate that conductance variations measured in spm experiments can not simultaneously reflect the spatial flow of electrons @xcite and the spatial form of @xmath111 @xcite .
different types of current patterns emerge when the size of the gnr is increased along the zig - zag direction , as shown in figures [ fig:7x](a ) - ( d ) . here
, we present the evolution of the current patterns with increasing @xmath112 in @xmath113 gnrs at @xmath96 .
the current patterns for the @xmath114 gnr ( which possesses a 7-fold degenerate state at @xmath92 , with 4 of these states possessing a non - zero wavefunction at gnr sites @xmath3 and @xmath4 ) is similar to that of the @xmath115 or @xmath74 gnrs , with the current propagating predominantly along the direction of the fermi velocity . with increasing @xmath112 ,
the current patterns become more complex , exhibit circulating current loops of various sizes [ see yellow arrows in figures [ fig:7x](d ) and ( h ) ] , and current backflow [ see yellow arrow in figure [ fig:7x](c ) ] .
we can identify an interesting relation between the number of degenerate states at @xmath92 , and the form of the current pattern : while the @xmath116 and @xmath117 gnrs possess only a single state at @xmath92 , and their spatial current patterns exhibit some similar subpatterns , the @xmath118 gnr possesses three degenerate states ( two of which possess a non - zero wavefunction at sites @xmath3 and @xmath4 ) , resulting in a current pattern that is qualitatively different from that of the @xmath116 and @xmath117 gnrs .
all of these current eigenmodes exhibit a different response to ( symmetric ) changes in the location of the leads , as shown in figures [ fig:7x](e)-(h ) . for the @xmath114 gnr , the current pattern simply deforms , but the current still propagates along the direction of the fermi velocity .
in contrast , for the @xmath116 , @xmath118 and @xmath117 gnrs , the current pattern undergoes a qualitative change when the leads are symmetrically displaced .
this reiterates the fact that charge transport through confined systems in the quantum regime is highly sensitive to boundary conditions , and the spatial form of the resulting coherent current eigenmode need not necessarily conform to expectations based on @xmath95 .
gnr with leads attached at different locations along the armchair edges.,width=529 ] the question naturally arises whether the same type of spatial current patterns emerge when the leads are attached to the armchair edges . to investigate this question , we consider a @xmath114 gnr [ the same gnr as shown in figure [ fig:7x](a ) ] and present in figure [ fig : arm ] the current patterns at @xmath96 for different locations of 4 leads attached to the armchair edges .
when the leads are attached to the center sites , as shown in figure [ fig : arm](a ) , the dominant contribution to the current flows along a line directly connecting the leads , rather than following the direction of the fermi velocity .
however , when the leads are displaced symmetrically from these high symmetry sites as shown in figures [ fig : arm](b ) and ( c ) , the current flows again primarily along the direction of the fermi velocity .
it is interesting to note that in all three cases [ figures [ fig : arm](a)-(c ) ] , the current exhibits two disjoint current paths , connecting either the left ( leads 1 and 3 ) or right ( leads 2 and 4 ) pair of leads . when one of these pairs of leads is removed ( and the spatial symmetry of the system is thus broken ) , as shown in figures [ fig : arm](d ) where only leads 2 and 4 remain , the current pattern becomes rather diffuse ( a very similar current pattern is found when only leads 2 and 3 remain ) .
this not only demonstrates that the symmetry of the armchair edges requires two leads to be attached to either armchair edge for a well - defined current pattern to emerge , but also that quantum interference effects between all four leads are crucial for creating the well - ordered current patterns shown in figures [ fig : arm](b ) and ( c ) .
gnrs possess two low energy states near the middle of the band at @xmath97 , i.e. , in close proximity to @xmath119 , that are delocalized along the zig - zag edge , and localized along the direction of the armchair edge @xcite .
the form of charge transport through these localized states in the wide - lead limit has recently attracted some attention @xcite , in particular in view of its possible application for dna sequencing @xcite . in figure
[ fig : edge](a ) , we present the local density of states for a @xmath120 gnr at @xmath121 , that clearly demonstrates the localized nature of these low energy states . demonstrating the existence of an edge state that is delocalized along the zig - zag edges , and decays exponentially along the armchair direction .
spatial current patterns for a voltage bias applied ( b ) between the armchair edges , and ( c ) between the zig - zag edges.,width=453 ] as a result , we expect that the charge transport at @xmath122 is highly anisotropic . to investigate this anisotropy , we consider the @xmath120 gnr with four leads attached to sites along their zig - zag edges , as shown in figures [ fig : edge](b ) and ( c ) .
when a voltage bias is applied between the armchair edges with @xmath122 , ( with chemical potential @xmath123 for leads 3 and 4 , and @xmath124 for leads 1 and 2 ) , the current flow is highly localized along the zig - zag edges , as shown in figure [ fig : edge](b ) , where the @xmath125 state is delocalized . in this case , the conductance of the @xmath125 state ( in the four - lead configuration ) is equal to the quantum of conductance ( similarly , the state at @xmath126 possesses a conductance given by the quantum of conductance ) .
this result holds for all values of @xmath100 or @xmath112 that we have considered .
moreover , for an infinitely large gnr in the armchair direction , i.e. , @xmath127 , one has a doubly degenerate state at @xmath128 ( one localized state at each of the zig - zag edges ) , and the conductance of the gnr at @xmath119 is as expected equal to twice the quantum of conductance .
note that the spatial current pattern in figure [ fig : edge](b ) is qualitatively different from that obtained when wide leads are attached to the armchair edges : in this case , the largest current density occurs in the center of the gnr @xcite and not along the zig - zag edges . on the other hand , when a voltage bias is applied between the zig - zag edges ( with chemical potential @xmath123 for leads 1 and 3 , and @xmath124 for leads 2 and 4 ) , the current is strongly suppressed [ see figure [ fig : edge](c ) ] , and scales as @xmath129 . since @xmath130 decreases exponentially with increasing length @xmath1 @xcite , we find that the current along the armchair direction also decreases exponentially with increasing length of the gnr , as expected from the localized nature of the @xmath131-states
. do quantum mechanical currents obey the principle of superposition ?
in particular , can spatial current patterns in gnrs that are attached to four leads , simply be obtained by superposing two - lead current patterns such as the ones discussed above ? to address this question we compare the spatial current patterns at @xmath96 in 4-lead and 2-lead configurations for two gnrs with different aspect ratios . in figure [ fig:4leads_1 ] , we present the current patterns for a @xmath132 gnr in 4-lead configurations with the leads attached symmetrically around the center of the gnr , as shown in figures [ fig:4leads_1](a ) - ( c ) , or with two leads attached , as shown in figures [ fig:4leads_1](d ) - ( f ) . in a @xmath132 gnr for ( a)-(c ) 4-lead and ( d ) -
( f ) 2-lead configurations.,width=453 ] since the gnr possesses 9 degenerate states at @xmath92 , we again find that the current follows predominantly the direction of the fermi velocity , thus propagating along the direction of the primitive lattice vector .
moreover , changing the separation between the leads simply deforms the current pattern , but does not introduce any qualitatively new spatial elements .
a comparison of the 4-lead current patterns in figures [ fig:4leads_1](a ) and ( b ) with the current patterns in a symmetric 2-lead configuration shown in figures [ fig:4leads_1](d ) and ( e ) , respectively , demonstrates that the 4-lead current pattern indeed represents a superposition ( i.e. , is the sum ) of two 2-lead current patterns .
moreover , the current flowing in the 4-lead configuration is twice that of the 2-lead configurations .
this implies that in the 4-lead configuration , there are no interference effects between the two superposing 2-lead current patterns , and that , in particular , no current flow occurs between the upper ( leads 1 and 3 ) or lower ( leads 2 and 4 ) pair of leads .
we also note that the 4-lead current pattern in figure [ fig:4leads_1](b ) is not the superposition of the 2-lead current pattern with just the upper two leads ( leads 1 and 3 ) in figure [ fig:4leads_1](f ) , with its respective counterpart . in this case
, the spatial current pattern is very diffuse , since the two leads can not be connected by a current path that propagates along @xmath133 , and the total current is only approximately 58.5% of that in the 2-lead configuration of figure [ fig:4leads_1](e ) .
gnr at @xmath96 for ( a)-(c ) 4-lead and ( d ) 2-lead configurations.,width=377 ] in contrast , a @xmath117 gnr possesses only a single state at @xmath92 , and the spatial form and complexity of the current pattern at @xmath96 changes qualitatively when the separation between the leads on each side of the gnr is varied , as shown in figures [ fig:4leads_2](a)-(c ) . when the leads are attached to the corners
[ see figure [ fig:4leads_2](a ) ] , the current flows along the edges and through the center of the gnr . at the same time , there are two large , circulating current loops in the upper and lower parts of the gnr , that do not contribute to the net current through the gnr .
in contrast , when the separation between the leads is reduced [ see figure [ fig:4leads_2](b ) ] , the current flows along the direction of @xmath133 , which directly connects the leads , while most of the gnr does not exhibit any charge flow at all . upon reducing the separation between leads even further
, the current pattern again becomes more complex [ see figure [ fig:4leads_2](c ) ] : the current now flows primarily along the outer perimeter , while connected to a larger circulating current loop in the center of the gnr . in figure
[ fig:4leads_2](d ) , we present the corresponding 2-lead current pattern ; the superposition of this pattern with its counterpart yields a spatial current pattern that is identical to that of figure [ fig:4leads_2](c ) .
however , the total current through the 4-lead and 2-lead configurations is the same since [ in contrast to the @xmath132 gnr ] there is only a single state that contributes to the charge transport at @xmath96 , thus limiting the conductance for both configurations to a single quantum of conductance . in other words ,
the spatial current in the 4-lead configuration is half the sum of the currents of the two 2-lead configurations .
the degeneracy of the @xmath92 state therefore does not only affect the spatial form of the current patterns in 2-lead configurations , but also whether the current patterns and total conductance in 4-lead configurations arises as a superposition of current patterns and conductances of 2-lead configurations .
the observation that localized states exist near zig - zag edges in gnrs raises the interesting question of the form of localized state in other gnr geometries @xcite . to investigate this question , we consider the diamond - shaped gnr shown in figure [ fig : diamond ] whose edges are only of the zig - zag type . in a diamond - shaped gnr .
( b ) spatial current pattern for @xmath96 , and ( c ) local density of states at @xmath92.,width=302 ] a plot of the local density of states at @xmath119 [ see figure [ fig : diamond](a ) ] reveals that a localized state exists , but that it is not confined to the zig - zag edges of the gnr [ as was the case for the gnr shown in figure [ fig : edge](a ) ] but to its two corners with acute angles .
moreover , while the gnr possesses three degenerate states at @xmath92 , we find that the current does not follow the direction of the fermi velocity ( which would require it to flow along the edges of the gnr ) , but rather flows through the gnr s center , as shown in figure [ fig : diamond](b ) .
finally , a plot of the local density of states at @xmath92 in figure [ fig : diamond](c ) , which is the largest in the gnr s center , again reveals no resemblance to the spatial current pattern , as previously discussed in the context of figure [ fig : x5 ] .
the ability to experimentally visualize spatial current patterns is crucial for understanding and manipulating transport at the nanoscopic or atomic scale . in mesoscopic systems , such as quantum point contacts @xcite , quantum rings @xcite and dna @xcite , imaging of spatial current paths
was successfully achieved using scanning probe microscopy ( spm ) @xcite .
the spatial resolution of spm , however , is insufficient to image spatial current patterns predicted to exist in nanoscopic systems @xcite which vary on the atomic scale .
two of us therefore recently proposed a novel method for the imaging of spatial current paths in nanoscopic systems based on scanning tunneling microscopy ( stm ) @xcite , a method that can resolve spatial current patterns on the scale of a lattice constant .
in addition , in the experimentally realized weak tunneling limit @xcite , this stm method only probes but essentially does not perturb the gnr s electronic structure , in contrast to spm . to demonstrate that this method for the spatial imaging of currents can also successfully be applied to gnrs , we consider as an example the current patterns in figures [ fig:7x](a ) and ( e ) for the @xmath114 gnr , and in figure [ fig : x5](e ) for the @xmath74 gnr . to visualize these spatial current patterns , we plot the spatial dependence of the current , @xmath134 , that flows from an stm tip ( held above the gnr at potential @xmath135 ) through the gnr into the left ( l ) or right ( r ) lead as a function of tip position @xmath136 ( for a more detailed description , see ref .
the tunneling of electrons from the stm tip into the gnr is described by the hamiltonian in equation ( [ eq : tunnel ] ) . here
, we set the bias between the leads to zero , i.e. , @xmath137 , and require that the bias difference between the tip and the system , @xmath138 be equal to @xmath139 , and that the bias midpoint @xmath140 be equal to @xmath73 ( here , @xmath73 and @xmath139 were used to generate the actual current patterns ) .
flowing from the stm tip into the left lead .
( d ) conductance @xmath141 as a function of bias midpoint @xmath73 for the @xmath114 gnr . here ,
we used @xmath142.,width=396 ] in figures [ fig : stm](a ) - ( c ) , we present the resulting stm image , i.e. , the spatial plots of @xmath143 , which correspond to the current patterns in figures [ fig:7x](a ) and ( e ) and in figure [ fig : x5](e ) , respectively .
a comparison of the actual current patterns with the stm images shows that the latter provide an accurate and atomically resolved image of the spatial current patterns .
this agreement holds even when the leads are not located at high symmetry positions , as shown in figures [ fig:7x](e ) and [ fig : stm](b ) , or when the size of the system increases , and the current pattern becomes repetitive , as in figures [ fig : x5](e ) and [ fig : stm](c ) .
moreover , we demonstrated in ref .
@xcite , that the form of the total conductance , @xmath141 , is an independent , experimentally verifiable , criterion for the success of the stm method in imaging spatial current patterns . in particular , we found that when @xmath141 is close to the maximal allowed conductance , and varies weakly around a given @xmath73 , the stm method successfully images the spatial current patterns at @xmath73 in the entire nanoscopic system . on the other hand ,
if @xmath141 is sharply peaked , the agreement breaks down in at least part of the network .
the @xmath114 gnr considered here satisfies this criterion , as follows from figure [ fig : stm](d ) , where we present its conductance @xmath141 : it is close to the quantum of conductance at @xmath144 ( see arrow ) where we image the current pattern and its width ( as a function of @xmath73 ) is larger than the distance between two peaks .
note that the zeros in the conductance @xmath141 correspond to zeroes of the real part of the retarded green s function for @xmath145 which occur near , but never exactly on , conductance resonances .
more generally , we find that the above criterion for @xmath141 is satisfied in nanoscopic networks which ( a ) possess a continuum of energy eigenstates through which transport can take place or ( b ) are strongly coupled to leads .
finally , we note in passing that understanding the magnitude and spatial paths of currents injected from the stm tip into the gnr , will likely also have relevance for dna sequencing where currents are induced locally by pulling a dna through a hole in a gnr @xcite .
for an @xmath74 gnr with a constriction ( shown in gray ) : ( a ) @xmath146 with mean free path @xmath147 ( b ) @xmath148 , @xmath149 , ( c ) @xmath150 , @xmath151 , ( d ) @xmath152 , @xmath153.,width=377 ] to study the effects of dephasing on the form of spatial current patterns , we consider the coupling of electrons to a local phonon mode @xcite , as described by the hamiltonian in equation ( [ eq : hamiltonian ] ) . as discussed in section [ sec : theory ] , we employ the high - temperature approximation , where the dephasing is controlled by a single parameter , @xmath154 [ see equation ( [ eq : sigma ] ) ] . in figure [ fig : dephasing ] we present the evolution of the spatial current pattern with increasing @xmath154 for an @xmath74 gnr that contains a constriction shown in gray . in the absence of dephasing ,
i.e. , for @xmath155 , [ see figure [ fig : dephasing](a ) ] , the current pattern is similar to that of the @xmath74 gnr without a constriction , as shown in figure [ fig : x5](e ) .
we note that while the constriction is located in a region of the gnr through which no current flows in figure [ fig : x5](e ) , it nevertheless leads to a change in the actual current pattern . as @xmath154 increases ,
the current pattern becomes more diffuse , and evolves smoothly from the ballistic limit [ see figure [ fig : dephasing](a ) ] , to that of a classical resistor network [ see figure[fig : dephasing](d ) ] .
moreover , with increasing @xmath154 , the current pattern remains well defined in the vicinity of the leads , but becomes more diffuse as one moves further away from the leads [ see regions indicated by yellow arrows in figure [ fig : dephasing](c ) ] .
this is expected since the diffuse current pattern arises from multiple scattering of the electrons off phonons , and the resulting dephasing . while required by symmetry
, it is nevertheless interesting to note that the effects of dephasing are reversed ( with the spatial current pattern again becoming more coherent ) as the current approaches the site where it exits the gnr .
this phenomenon of current patterns exhibiting coherent ( well - defined ) and incoherent ( diffusive ) spatial regions occurs when the mean - free path is considerably shorter than the size of the gnr [ in figure [ fig : dephasing](c ) , one has @xmath151 ] , but larger than a few lattice spacings . as the mean - free path becomes shorter , the region around the leads where the current forms a well defined , coherent pattern shrinks and eventually vanishes
[ see figure [ fig : dephasing](d ) ] . in the preceding sections
, we neglected the effects of a possible next - nearest neighbor hopping , @xmath12 , on the form of the electronic excitation spectrum and the resulting conductance and spatial currents patterns in gnrs .
the question naturally arises to what extent a non - zero @xmath12 , with estimates ranging from @xmath157 to @xmath158 @xcite , qualitatively affects the results presented above . in macroscopic graphene lattices ,
the existence of a non - zero @xmath12 is often neglected , not only because of the large uncertainty in its value , but also because its effects on the electronic excitation spectrum are rather trivial . in particular ,
the energy dispersion for a non - zero @xmath12 is given by @xcite @xmath159 where @xmath160 as a result , all momentum states located at an energy @xmath161 for @xmath162 are simply shifted to a new energy @xmath163\ ] ] for @xmath164 .
this , in particular , implies that the degeneracy of states is not lifted by a non - zero @xmath12 . for finite - size gnrs , however , the situation is different since a non - zero @xmath12 lifts the degeneracy of the states which for @xmath165 are located at @xmath92 states .
the extent of the energy splitting between these states is non - universal and depends not only on the aspect ratio of the gnr , but also on its overall size , since , according to eq.([eq : ektp ] ) , the splitting vanishes in the limit of infinitely large graphene layers where degeneracy is restored . when @xmath166 .
if @xmath24 are chosen to bracket all three conductance resonances , well - ordered current patterns are produced for the ( b ) @xmath115 gnr at @xmath167 with @xmath168 , @xmath169 , and @xmath170 , and ( c ) @xmath171 gnr at @xmath172 with @xmath173 , @xmath174 , and @xmath175 .
( d ) the ldos for the edge states at energy @xmath176 . , width=566 ] in figure [ fig : nnn](a ) , we present a conductance scan for a @xmath115 gnr with @xmath177 over an energy range that contains all three states ( and only those ) that are located at @xmath92 for @xmath166 , and possess a non - zero wavefunction at @xmath3 and @xmath4 . choosing a left and right chemical potential that brackets these three states [ see vertical red lines in figure [ fig : nnn](a ) ] , we find that the resulting spatial current pattern shown in figure [ fig : nnn](b ) still agrees very well with that obtained for the case @xmath166 , as follows from a comparison of figures [ fig : nnn](b ) and [ fig : x5](b ) .
the same conclusion also holds for larger gnrs and different values of @xmath12 , as follows from a comparison of the current pattern for a @xmath74 gnr with @xmath178 shown in figure [ fig : nnn](c ) with that for @xmath165 shown in figure [ fig : x5](e ) .
this implies that for both cases , @xmath166 and @xmath179 , the same ( well - ordered ) spatial current pattern occurs if the left and right chemical potentials are chosen such that they bracket all states which for @xmath166 occur at @xmath92 .
finally , we found that a non - zero @xmath12 shifts the energy of the state localized along the zig - zag edge near @xmath180 for @xmath166 , but does not change its localized nature or spatial structure . in figure
[ fig : nnn](d ) we present the local density of states for @xmath181 at @xmath182 , which demonstrates that this state remains localized along the zig - zag edge .
in this article , we investigated the quantum nature of local charge transport in graphene nanoribbons , as reflected in the form of spatial current patterns . by using the non - equilibrium keldysh green s function formalism @xcite , we demonstrated that the form of spatial current patterns is determined by the interplay between the gnrs geometry , size and aspect ratio , by the location and number of leads , and the presence of dephasing .
in particular , we identified a crucial relation between the spatial form of current patterns , and the number of degenerate states participating in the transport .
this insight , in principle , provides us with the opportunity to predict and custom - design spatial current patterns in gnrs .
furthermore , we showed that the number of degenerate states participating in charge transport also determines whether current patterns and conductances in gnrs with 4-lead configurations can be considered as arising from the superposition of current patterns and conductances in 2-lead configurations .
in addition , we showed that spatial current patterns in gnr can be spatially imaged with atomic resolution using scanning tunneling microscopy , allowing us to gain unique insight into the nature of local charge transport .
we also demonstrated how spatial current patterns evolve with increasing dephasing from the ballistic limit , where they are coherent and spatially well defined , to the classical limit , where they are incoherent and spatially diffuse .
finally , we showed that our conclusions remain valid in the presence of a realistic next - nearest neighbor hooping element .
these results represent an important first step towards understanding and manipulating charge transport at the atomic level in gnrs , which are a necessary requirement for the further development of graphene based nanoelectronics and dna sequencing .
+ * acknowledgements * we would like to thank d. goldhaber - gordon and h. manoharan for stimulating discussions .
this work is supported by the u.s .
department of energy , office of basic energy sciences , division of materials sciences and engineering under award no .
de - fg02 - 05er46225 ( d.k.m ) and by a u.s .
department of education gaann fellowship and mrsec nsf dmr-0820054 ( t.c . ) .
< num > munoz - rojas f , jacob d , fernandez - rossier j and palacios j j 2006 _ phys .
_ b * 74 * 195417 cresti a , grosso g and parravicini g p 2007 _ phys .
_ b * 76 * 205433 zarbo l p and nikolic b k 2007 _ eur .
lett . _ * 80 * 47001 li z , qian h , wu j , gu b - l and duan w 2008 _ phys .
lett . _ * 100 * 206802 maiti s k 2009 _ solid state commun . _ * 149 * 973 gonzalez j w , santos h , pacheco m , chico l and brey l 2010 _ phys
_ b * 81 * 195406 takaki h and kobayashi n 2010 _ japan j. appl .
phys . _ * 49 * 08lb04 takaki h and kobayashi n 2011 _ physica _ e * 43 * 711 das sarma s , adam s , hwang e h and rossi e 2011 _ rev . mod . phys .
_ * 83 * 407 lu c - h , yang h - h , zhu c - l , chen x and chen g - n 2009 _ angew . chem .
_ * 48 * 4785 pumera m 2011 _ mater . today
_ * 14 * 308 chowdhury r , adhikari s , rees p , wilks s p and scarpa f 2011 _ phys .
_ b * 83 * 045401 ; chowdhury r , scarpa f and adhikari s 2012 _ j. appl
. phys . _ * 112 * 104905 schneider g f , kowalczyk s w , calado v e , pandraud g , zandbergen h w , vandersypen l m k and dekker c 2010 _ nano lett . _
* 10 * 3163 min s k , kim w y , cho y and kim k s 2011 _ nat .
nanotechnol . _ * 6 * 162 venkatesan b m and bashir r 2011 _ nat .
nanotechnol . _ * 6 * 615 saha k k , drndic m and nikolic b k 2012 _ nano lett . _
* 12 * 50 li x , wang x , zhang l , lee s and dai h 2008 _ science _ * 319 * 1229 han m y , ozyilmaz o b , zhang y and kim p 2007 _ phys .
lett . _ * 98 * 206805 bolotin k i _ et al _ 2008 _ solid state commun . _ * 146 * 351 ; du x , skachk i , barker a and andrei e y 2008 _ nat .
nanotechnol . _
* 3 * 491 gomes k k_et al _ 2012 _ nature _ * 483 * 306 kumar s b _ et al _ 2010 _ j. phys . : condens . matter _ * 22 * 375303 for more on imaging spatial current flow using spm , see : eriksson m a , beck r g , topinka m , katine j a , westervelt r m , campman k l and gossard a c 1996 _ appl . phys . lett . _ * 69 * 671 ; topinka m a , leroy b j , westervelt r m , shaw s e j , fleischmann r , heller e j , maranowski k d and gossard a c 2001 _ nature _ * 410 * 183 ; aidala k e , parrott r e , kramer t , heller e j , westervelt r m , hanson m p and gossard a c 2007 _ nature physics _ * 3 * 464 hackens b , martins f , ouisse t , sellier h , bollaert s , wallart x , cappy a , chevrier j , bayot v and huant s 2006 _ nature physics _ * 2 * 826 ; hackens b , martins f , faniel s , dutu c a , sellier h , huant s , pala m , desplanque
l , wallart x and bayot v 2010 _ nature commun . _ * 1 * leroy b j 2003 _ j. phys . : condens
. matter _ * 15 * r1835 ; park j y , maier s , hendriksen b and salmeron m 2010 _ mater . today _ * 13 * 37 ; sellier h , hackens b , pala m g , martins f , baltazar s , wallart x , desplanque l , bayot v and huant s 2011 _ semicond .
technol . _
* 26 * 064008 lent c s
1990 _ appl .
_ * 57 * 1678 ; berggren k f , besev c and ji z - l 1992 _ phys .
_ * t42 * 141 ; nonoyama s and oguri a 1998 _ phys .
_ b * 57 * 8797 ; laux s e , kumar a and fischetti m v 2004 _ j. appl .
phys . _ * 95 * 5545 | using the non - equilibrium keldysh green s function formalism , we investigate the local , non - equilibrium charge transport in graphene nanoribbons ( gnrs ) .
in particular , we demonstrate that the spatial current patterns associated with discrete transmission resonances sensitively depend on the gnrs geometry , size , and aspect ratio , the location and number of leads , and the presence of dephasing .
we identify a relation between the spatial form of the current patterns , and the number of degenerate energy states participating in the charge transport .
furthermore , we demonstrate a principle of superposition for the conductance and spatial current patterns in multiple - lead configurations .
we demonstrate that scanning tunneling microscopy ( stm ) can be employed to image spatial current paths in gnr with atomic resolution , providing important insight into the form of local charge transport .
finally , we investigate the effects of dephasing on the spatial current patterns , and show that with decreasing dephasing time , the current patterns evolve smoothly from those of a ballistic quantum network to those of classical resistor network . |
the exclusive electroproduction of mesons on the nucleon is an important tool to better understand the nucleon structure and , more generally , the transition between the low energy hadronic and high energy partonic domains of the quantum chromodynamics ( qcd ) theory .
the cebaf ( `` continous electron beam accelerator facility '' ) of the jefferson laboratory ( jlab ) at newport news ( usa ) with its high intensity electron beam and its large acceptance spectrometer clas @xcite ( cebaf large acceptance spectrometer ) offers a great opportunity to study the exclusive electroproduction of mesons , in particular those that have multi - particle decays . + this proceeding presents the results for the exclusive electroproduction of the vector mesons @xmath1 , @xmath2 and @xmath3 on the proton at clas and provides a first ( preliminary ) look at the @xmath4 channel .
these results come mainly from two experiments : the e1 - 6 and the e1-dvcs experiments . for e1 - 6
, the data were collected between october 2001 and january 2002 .
the beam energy was 5.754 gev and the integrated luminosity was 28.5 fb@xmath5 .
the e1-dvcs data were collected between march and may 2005 .
the beam energy was 5.75 gev and the integrated luminosity was @xmath6 fb@xmath5 .
the cross sections for the @xmath1 , @xmath2 and @xmath3 channels are published in refs .
@xcite . for the @xmath4 channel , it is the first ever measurement of its cross section .
the presented results are still preliminary and more details can be found in ref . @xcite .
the final state that is analyzed is @xmath8 . the final state particles @xmath9 , @xmath10 and @xmath11 were detected in clas and the exclusive process was identified by selecting a missing @xmath12 with the missing mass technique .
[ missigep ] shows the distributions of @xmath13 $ ] for some @xmath14 bins .
one sees clearly the @xmath1 peak sitting on top of a background of a non - resonant two - pion continuum .
one can also distinguish two bumps at masses around 980 mev and 1270 mev corresponding , respectively , to the scalar @xmath15 and tensor @xmath16 mesons .
+ after weighting each event with the appropriate acceptance , we have used four contributions to fit these spectra : three skewed breit - wigner distributions to describe the @xmath1 , @xmath17 and @xmath18 resonant structures and a parametrisation of the @xmath19 non - resonant continuum determined from simulations ( mostly @xmath20 phase space ) . after the background subtraction procedure , total and differential cross sections of the reaction @xmath21 were extracted , by calculating for each @xmath14 bin the area of the @xmath1 breit - wigner .
$ ] for some @xmath14 bins .
@xcite for the kinematics of each bin.,width=340 ] a separation of the longitudinal cross sections from the transverse ones was performed by analysing the pion decay angles of the @xmath1 and relying on the schc ( @xmath22-channel helicity conservation ) which was checked experimentally .
the final state that is analyzed is @xmath24 .
the final state particles @xmath9 , @xmath10 and @xmath11 were detected in clas and the exclusive process was identified by the missing mass @xmath25 $ ] .
[ omegaphi ] ( left ) shows this acceptance - corrected missing mass distribution for a particular ( @xmath26 ) bin .
the @xmath2 peak is fitted with a skewed gaussian shape ( taking into account the experimental resolution and radiative tail ) and the background with a second - order polynomial .
$ ] distribution for a particular ( @xmath26 ) bin .
right : invariant mass @xmath27 for the whole data set.,title="fig:",width=188,height=113]$ ] distribution for a particular ( @xmath26 ) bin .
right : invariant mass @xmath27 for the whole data set.,title="fig:",width=188,height=113 ] the final state that is analyzed is @xmath29 .
the final state particles @xmath9 , @xmath10 and @xmath30 were detected in clas and the exclusive process was identified by selecting a missing @xmath31 with the missing mass technique .
[ omegaphi ] ( right ) shows the invariant mass @xmath27 for the entire data set with a clear @xmath3 meson peak .
this distribution is fitted with a gaussian plus a polynomial function for the background .
after weighting each event with the acceptance , the background was subtracted and the cross section of the exclusive electroproduction of @xmath3 was extracted .
the final state that is analyzed is @xmath33 .
the final state particles @xmath9 , @xmath11 and @xmath34 were detected in clas and the exclusive process was identified by selecting a missing neutron with the missing mass technique .
[ fitq2xb ] shows the acceptance - corrected invariant mass @xmath35 distributions for each @xmath14 bin .
one sees clearly the @xmath36 peaks around 775 mev .
they are however sitting on top of a non - negligeable non resonant two - pion background .
we have used two contributions to fit these spectra : a skewed breit - wigner distribution to describe the resonant structure of the @xmath4 and a parametrization of the non - resonant two - pion continuum determined from simulations ( mostly @xmath33 phase space ) . in order to extract the total cross section of the reaction @xmath37 , we have calculated for each
@xmath14 bin the area of the breit - wigner ( green curve in fig .
[ fitq2xb ] ) , the distributions being already normalized and corrected by the acceptance .
acceptance - corrected distributions , showing fits for the background subtraction . in black : experimental data ; in green : breit - wigner of @xmath4 ; in purple : @xmath35 projection of the the non - resonant continuum @xmath38 reaction ; in red : total fit result . in this proceeding , units are arbitrary for the vertical scale.,width=453 ]
we now compare the extracted total and differential cross sections of exclusive electroproduction for these vector mesons with two theoretical approaches : on the one hand , the hadronic approach based on regge theory and meson trajectory exchanges in the @xmath0-channel and on the other hand , the partonic approach based on the handbag diagram and gpds .
these two approaches are illustrated in fig .
[ reggegpds ]
. this comparison will be useful in order to better understand the domain of validity of these two approaches and constrain their inputs .
-channel exchange ( left ) and of the handbag diagram ( right ) for exclusive vector meson electroproduction.,title="fig:",height=113]-channel exchange ( left ) and of the handbag diagram ( right ) for exclusive vector meson electroproduction.,title="fig:",height=113 ] the regge approach is most appropriate above the resonance region and at forward angles where the cross section is the largest .
it consists of exchanges of meson `` trajectories '' in the @xmath0-channel . in the following
, we will use the `` jml '' acronym to refer to the specific model developed by j .-
laget and collaborators @xcite . + for the @xmath1 channel , the model consists of @xmath0-channel exchanges of @xmath39 , @xmath18 and pomeron trajectories . for the @xmath2 channel , the exchange particles are @xmath34 and pomeron trajectories . for the @xmath3 channel ,
the only exchange particle is the pomeron trajectory . for the @xmath4 channel ,
the exchange particles are the @xmath11 and the @xmath4 trajectories .
the free parameters are the coupling constants at the hadronic vertices ( most of them being well constrained ) and the mass scales of the electromagnetic form factors
[ theoryrhzero ] shows the total longitudinal cross section @xmath40 as a function of @xmath41 for fixed @xmath42 .
the results of the jml calculation is shown with the dot - dashed curve .
the jml model is able to successfully reproduce the cross sections for almost all of our @xmath43 range . at fixed @xmath42 , for the reaction @xmath44 .
the dashed curve shows the result of the gk calculation and the thin solid curve shows the result of the vgg calculation .
the thick solid curve is the vgg calculation with the addition of the new @xmath0-channel meson exchange term .
the dot - dashed curve shows the results of the regge jml calculation .
the references of the presented data can be found in ref .
@xcite.,width=415 ] it can also reproduce successfully the main trends of the @xmath41 , @xmath42 and @xmath0 dependences of the total and differential cross sections of the reactions @xmath45 , @xmath46 and @xmath47 as presented in refs . @xcite . the formalism of gpds is valid in the so - called bjorken regime , i.e. @xmath48 with @xmath49 finite .
it was proven @xcite that the dominant process for exclusive meson electroproduction , in the bjorken limit , is given by the so - called handbag diagram represented in fig .
[ reggegpds ] .
the handbag diagram is based on the notion of factorization in leading - order / leading - twist pqcd between a hard scattering process , exactly calculable in pqcd , and a nonperturbative nucleon structure part that is parametrized by the gpds . for mesons ,
the factorization of the handbag diagram is only valid for the longitudinal part of the cross section .
for electroproduction of @xmath1 and @xmath2 we are sensitive to the sum of the quark handbag diagram and the gluon handbag diagram , while for the @xmath4 and the @xmath3 channels we are sensitive to , respectively , only the quark handbag diagram and the gluon handbag diagram .
+ in the following we discuss the two particular gk @xcite and vgg @xcite gpd - based calculations that provide quantitative results for the longitudinal part of the exclusive meson cross sections . in fig .
[ theoryrhzero ] , the dashed line shows the results of the gk model , while the thin solid line shows the result of the vgg model .
we see that they give a good description of the high and intermediate @xmath41 region , down to @xmath50 gev . at high @xmath41
the slow rise of the cross section is due to the gluon and sea contributions , while the valence quarks contribute only at small @xmath41 . at lower @xmath41 values , where the new clas data lie ,
both the gk and vgg models fail to reproduce the data .
this discrepancy can reach an order of magnitude at the lowest @xmath41 values .
the trend of these particular gpd calculations is to decrease as @xmath41 decreases , whereas the data increase .
the same behavior was observed , in the low @xmath41 region , for the exclusive electroproduction of the @xmath4 as one sees in fig .
[ theoryrhoplus ] .
the results of the calculations of the gk and vgg models are shown , respectively , with the red and the blue curves . +
an attempt to reconcile the gpd calculation with the low @xmath41 @xmath1 cross sections is presented in ref.@xcite . through a toy - model , @xmath0-channel meson exchanges
are included in the gpds and the result of this calculation ( actually a fit ) is illustrated by the thick blue curve in fig .
[ theoryrhzero ] . + fig .
[ phigpds ] shows the longitudinal cross section of @xmath3 electroproduction compared with the calculation of the gk model .
one can see that the model reproduces well the data for all the presented @xmath41 range : for low @xmath41 data where the clas data lie and for higher @xmath41 .
as mentionned before , for the @xmath3 channel , we have only the contribution of the handbag diagram of gluons and this comparison shows that the gpd approach can work well for gluons . + at fixed @xmath42 , for the reaction @xmath51 .
the red and the blue curves are the results of the gk and vgg models .
units are arbitrary on the y axis.,width=377,height=264 ] , for the reaction @xmath52 .
the references of the presented data can be found in ref .
the black curve is the result of the gk model.,width=302,height=188 ] from the results of the three mesons channels @xmath1 , @xmath4 and @xmath3 , one can conclude that the gpd approach seems to work for gluons and sea quarks and largely fail for valence quarks .
many theorical efforts are now in progress to understand this transition between the small @xmath41 region and the high @xmath41 region . + for the @xmath2 channel , schc was found to be violated by a significant amount , thus preventing a longitudinal / transverse cross section separation .
we therefore do not discuss here its comparison with gpd calculations .
[ tslopes ] shows the slope of the differential cross section @xmath53 for the @xmath1 , @xmath2 , @xmath3 and @xmath4 channels as a function of @xmath41 ( on the top part ) and as a function of @xmath42 ( in the bottom part ) .
one can see the same trends of this slope @xmath54 for all mesons channels , which can be interpreted in simple and intuitive terms in the following way : * @xmath54 increases with @xmath41 : the size of the nucleon increases as one probes the high @xmath41 values ( i.e. the sea quarks ) , which could mean that the sea quarks tend to extend to the periphery of the nucleon .
* @xmath54 decreases with @xmath42 : as we go to large @xmath42 , the resolution of the probe increases and we tend to see smaller and smaller objects . as a function of @xmath41 ( on the top ) and as a function of @xmath42 ( on the bottom ) for the @xmath1 , @xmath2 , @xmath3 and @xmath4 channels.,width=453 ]
using the clas detector at jlab , we have collected the largest ever set of data for the exclusive electroproduction of vector mesons on the proton in the valence region .
we have presented the published results for @xmath1 , @xmath2 and @xmath3 and given a first look at the @xmath4 ( preliminary ) cross sections .
these data can be interpreted with two approaches : * hadronic approach : the jml model describes well most of the features of the @xmath1 , @xmath2 , @xmath3 and @xmath4 cross sections up to @xmath55 gev@xmath56 . *
partonic approach : gpd models describe well the data for @xmath57 gev but largely fail for @xmath58 gev .
a comparison between the features of the different mesons is in progress : we find for instance the same trends of the variation of the @xmath0 slope as a function of @xmath41 and as a function of @xmath42 for all @xmath1 , @xmath2 , @xmath3 and @xmath4 channels .
99 b.a . mecking et al .
instr . meth .
a * 503 * , ( 2003 ) 513 .
s. morrow et al . , eur .
phys . j. a * 39 * , ( 2009 ) 5 .
l. morand et al . , eur .
phys . j. a * 24 * ( 2005 ) 445 .
j. p. santoro et al . , phys .
c * 78 * ( 2008 ) 025210 .
yao et al .
, j. phys . lett .
g 33 ( 2006 ) 1 .
a. fradi , thesis univ .
paris - sud at orsay , 2009 .
cassel et al . , phys .
d * 24 * ( 1981 ) 2787 . j .-
laget , phys .
b * 489 * ( 2000 ) 313 . ;
+ f. cano and j .- m .
laget , phys .
d * 65 * ( 2002 ) 074022 . ; + j. m. laget , phys .
d * 70 * ( 2004 ) 054023 . ; + f. cano and j. m. laget , phys .
b * 551 * ( 2003 ) 317 .
collins , l. frankfurt and m. strikman , phys.rev.d * 56 * ( 1997 ) 2982 .
goloskokov and p. kroll , eur .
* 42 * ( 2005 ) 281 ; eur.phys .
j.c * 50 * ( 2007 ) 829 .
m. vanderhaeghen , p.a.m .
guichon , m. guidal , phys .
d * 60 * , 094017 ( 1999 ) .
m. guidal and s. morrow , _ proceeding of the international workshop `` exclusive reactions at high momentum transfer '' , jefferson laboratory , newport - news , virginia , usa , may 21 - 24 2007 _ , world scientific , arxiv:0711.3743 ( hep - ph ) . | we present the results of exclusive electroproduction of vector mesons on the proton at clas .
we discuss the interpretation of these cross sections in terms of @xmath0-channel reggeon exchanges and in terms of generalized parton distributions ( gpds ) formalism . |
once the possibility of building a negative - index metamaterial ( nim ) was proven in the microwave regime @xcite , extraordinary effort has been made to obtain analogous behaviours for increasingly higher frequencies up to the visible range of the electromagnetic spectrum @xcite .
the major obstacle found when trying to translate ideas to higher frequencies was how to achieve a strong diamagnetic response in the systems designed .
many attempts to tackle the problem were simple miniaturizations of the canonical designs employed in those first metamaterials operating in microwaves . as an example ,
the split - ring configurations , or slight variations of it , were reduced up to the nanoscale to obtain such a magnetic response .
apart from the fundamental limitations inherited from those designs , e.g. anisotropy , and the increasingly complexity in the fabrication procedures , several drawbacks have been found in this process , some of them being consequence of the different behaviour of metals when excited with higher frequency waves @xcite .
precisely this different behaviour also inspired some authors to search for different configurations intended to exploit the plasmonic response to obtain artificial magnetism .
in many of them artificial magnetism is due to coupling between different plasmonic structures , their drawback thus being the high losses within metallic parts @xcite . in most cases , the proposed designs are restricted to operate under certain polarization and incidence conditions @xcite-@xcite , or are not truly three - dimensional materials @xcite .
moreover , in many cases , the behaviour of the proposed designs stems from coupling between the different constituents , thus making particular arrangements necessary . as a consequence
, spatial dispersion effects often appear due to propagation of waves in the lattice . lately
, some approaches based on plasmonic waveguides supporting negative - index modes have pushed the frequencies in which left - handed ( lh ) behaviour is obtained well within the visible spectrum .
however , propagation inside such material would be limited to the propagation length of the plasmon in the waveguide , thus making the design a single layer device , lacking truly three - dimensionality @xcite .
in addition , various attempts have been made towards nim by exploiting magnetic resonances occurring in structures made of high permittivity materials @xcite-@xcite .
some of them combine these structures with secondary structures providing the electrical response @xcite-@xcite or embed them in a metallic host medium @xcite , thus having inherent high losses . here
we report a design that tackle many of the previously mentioned limitations . a totally three - dimensional isotropic negative - index metamaterial operating at optical frequencies , and whose response is due to every isolated meta - atom .
therefore , no particular arrangement of them is needed .
we study the possiblity of using a spherical core - shell configuration to obtain with one single structure both electric and magnetic responses .
the core , being metallic , is responsible for the electrical response , while the shell , made of a high permittivity dielectric , provides the strong diamagnetic response .
an extension of mie theory is exploited to rigorously determine the scattering properties and resonances of the whole spherical core - shell configuration , which essentially determine the effective material properties @xcite .
calculations for core - shell structures built up with realistic materials ( ag@si and ag@ge ) demonstrate the possibility to obtain nim operating at 1.2 @xmath1m-1.55@xmath1 m . since both responses are attained directly from one single constituent , no particular arrangement of the inclusions is needed . in order to extract the effective parameters of the metamaterial
, we will assume both a random and a simple cubic distribution .
in the case of a random distribution lorentz - lorenz theory is applied , leading to simultaneously negative effective permeability and permittivity for several filling fractions . in the case of a simple cubic lattice arrangement ,
finite - element - method is applied to carry out numerical simulations , which fully account for interaction between the periodically arranged constituents .
effective material constants are then extracted through standard s - parameter @xcite retrieval procedure , and tested to fulfill causality and passivity , thus confirming a true double - negative index . due to the spherical symmetry of the constituents
, the metamaterial response will be essentially isotropic and polarization independent .
let us examine the scattering of a plane electromagnetic wave ( wavelength @xmath2 ) from a spherical core - shell particle without any approximation , which indeed can be done analytically as an extension of mie theory , obtained first by aden and kerker @xcite .
figure [ fig1](a ) depicts the geometry of one of the basic constituents or `` meta - atoms '' : a high permittivity ( @xmath3 ) dielectric shell is considered with outer radius @xmath4 and thickness @xmath5 , @xmath6 and @xmath7 being the dielectric constant of the core and the surrounding medium , respectively .
the scattering and extinction efficiencies can be expressed in terms of the material and geometrical parameters through the scattering coefficients @xmath8 and @xmath9 ( which represent , respectively , the different electric and magnetic multipolar contributions ) as : @xmath10 where @xmath11 .
the mentioned scattering coefficients can be written in terms of the the spherical bessel functions of the first ( @xmath12 ) and second ( @xmath13 ) class and depend on @xmath14 , @xmath15 , @xmath16 and @xmath17 .
@xmath18 denotes the real part .
their explicit form is : @xmath19-n\psi'_l(y)[\psi_l(ny)-a_l\chi_l(ny)]}{\xi_l(y)[\psi'_l(ny)-a_l\chi'_l(ny)]-n\xi'_l(y)[\psi_l(ny)-a_l\chi_l(ny)]}\ ] ] @xmath20-\psi'_l(y)[\psi_l(ny)-b_l\chi_l(ny)]}{n\xi_l(y)[\psi'_l(ny)-b_l\chi'_l(ny)]-\xi'_l(y)[\psi_l(ny)-b_l\chi_l(ny)]}.\ ] ] the ricatti - bessel functions introduced are @xmath21 , @xmath22 and @xmath23 , where @xmath24 is the spherical hankel function of the first class .
the coefficients @xmath25 and @xmath26 are : @xmath27 @xmath28 where @xmath29 , @xmath30 and @xmath31 .
it is inmediate to realize that all information about material and geometrical properties of the core is contained in these @xmath25 and @xmath26 coefficients . to better understand the physics behind , we first characterize the magnetic resonance of a nanoshell for a real dielectric material with high refractive index .
we choose silicon , si , the refractive index of which can be considered constant , @xmath32 , and lossless , within the near - ir range @xmath33 . in figure [ fig1](b )
the contribution is plotted to the total scattering efficiency of the magnetic dipolar term ( @xmath34 ) as a function of @xmath16 and the incidence wavelength .
it is the dominant contribution , as can be seen in figure [ fig1](c ) for the specific case of @xmath35 nm .
a resonance can be clearly seen , the wavelength of which corresponds to that of the si compact sphere @xcite when @xmath36 and starts to redshift when the shell thickness is thin enough ( @xmath37 ) and decreases .
thus the magnetic resonance of the si nanoshell can be tuned within a certain range of wavelengths larger than that of the si sphere resonance .
the behaviour of the em fields at the magnetic resonance has been calculated by full - wave ( finite - element method ) numerical simulations @xcite .
the result is depicted in figure [ fig1](d ) and ( e ) .
it can be observed that the electric displacement field rotates in planes parallel to the equator , thus inducing a strong magnetic moment .
this pattern clearly shows that the electrical displacement current @xmath38 is strongly confined inside the shell , and rotates along the @xmath39 direction around the incident h - field .
such behaviour reveals that the resonance is qualitatively quite similar to the first resonance of a dielectric ring , already reported @xcite , following an lc model with an inductance @xmath40 associated to the circulation of the displacement current , and a capacitance @xmath41 associated to the electric energy confined in the shell . at resonance , a strong magnetic moment along the direction of the incident h - field
is generated in the same qualitative way as in the dielectric ring analyzed @xcite .
nm as a function of the inner radius ( @xmath16 ) and wavelength of the incident light .
( c ) total scattering efficiency , together with the dipolar electric ( @xmath42 ) and magnetic ( @xmath34 ) contributions to the scattering efficiency for inner radius @xmath35 nm ( indicated in ( b ) by a dashed white line ) .
( d - e ) near - field plots at the magnetic resonance for the si shell of ( c ) .
( d ) norm of electric field .
( e ) out - of - plane component ( only non - zero component of the incident magnetic field ) of @xmath43 , together with the electric displacement field in white arrows . ]
next we analyze the mutual influence of a metallic core and a dielectric shell . with regard to the plasmon resonance of the metallic core ( i.e. , collective oscillations of conduction electrons )
, it is well known that for small spheres ( @xmath44 ) of dielectric permittivity @xmath45 , embedded in a medium with dielectric constant @xmath7 , the induced dipole moment is resonant at the frequency @xmath46 such that @xmath47 @xcite . therefore , when the metallic sphere is coated with a thick dielectric layer of permittivity @xmath3 , the quantity driving the resonance condition is not @xmath6 itself but , instead , the ratio between the permittivities of the core and the coating .
that is , for a small metal core , the resonance occurs when @xmath48 is fulfilled , thus redshifting the localized surface - plasmon resonance ( lspr ) with respect to that in vacuum .
the question now arises as to whether or not the shell magnetic resonance is preserved when the core is metallic .
since the confinement and rotation of the electric field inside the coating is directly related to the jump conditions for the normal component of the field between the shell and the surrounding medium @xcite , and since the metal is expected to avoid penetration of the field inside , no significant changes of the magnetic resonance behaviour are expected at least for relatively small metallic cores .
this will be revealed below , where we plot the scattering efficiency together with the electric and magnetic dipolar contributions .
such behaviour makes possible to still predict the appearance of the magnetic resonance even if a metallic core is present .
furthermore , the electric and magnetic responses of a metal - dielectric core - shell nanosphere can be tuned as to make both resonances coincide .
let us take silver ( ag ) as the material to build up the core and si again as a high permittivity dielectric for the shell .
complex ag permittivity values are taken from tabulated data @xcite .
our goal is to obtain simultaneously both resonances in the near infrared ( ir ) range of the electromagnetic spectrum .
the lspr will be redshifted to that wavelength such that @xmath49 .
this happens at a wavelength about 720 nm .
nevertheless , we have to keep in mind that , strictly , this would only be a good approximation for very small metallic particles with relatively thick coatings .
since we want to push the electric resonance to the ir to make it coincide with a magnetic one stemming from the si shell , we can take increasingly larger radius for the core .
when the size of a metallic particle is increased , there is a redshift in the resonance wavelength that can be explained in terms of depolarization effects @xcite
. nm as a function of the inner radius ( @xmath16 ) and wavelength of the incident light .
( d ) extinction and scattering efficiencies , together with the dipolar electric ( @xmath42 ) and magnetic ( @xmath34 ) contributions to the scattering efficiency for inner radius @xmath50 nm ( indicated in ( c ) by a dashed white line ) . (
e - g ) dipolar electric contribution ( e ) , dipolar magnetic contribution ( f ) , and total scattering efficiency ( g ) of a ag@ge core - shell nanosphere with outer radius @xmath51 nm as a function of the inner radius ( @xmath16 ) and wavelength of the incident light .
( h ) extinction and scattering efficiencies , together with the dipolar electric ( @xmath42 ) and magnetic ( @xmath34 ) contributions to the scattering efficiency for inner radius @xmath52 nm ( indicated in ( g ) by a dashed white line ) . ]
figures [ fig2](a)-(c ) depicts both , the electric dipolar and the magnetic dipolar contributions to the scattering efficiency , together with the total scattering efficiency for a ag@si core - shell system of outer radius @xmath53 nm as a function of @xmath16 and the incidence wavelength .
it can be clearly seen an overlap between the electric and magnetic resonances .
it happens within 1150 nm and 1300 nm and for inner radius between @xmath54 nm .
we have explicitly plotted the case @xmath50 nm ( figure [ fig2](d ) ) .
an interesting feature that can be observed is that the magnetic resonance disappears as the inner radius increases .
it can be explained by the fact that the electric displacement field in the core rotates in the opposite direction as it does in the shell , thus reducing the total magnetic moment generated .
another effect can be seen in the electric dipole contribution . as the inner radius increases the resonance broadens and redshifts as expected .
interestingly , this behaviour changes when the thickness is comparable with @xmath16 , and the resonance starts to blueshift .
this effect can be attributed to the fact that , as the dielectric shell becomes thinner , the resonance wavelength tends to that of a sphere without coating , thus blueshifting . finally , it is important to note that the specific wavelengths at which the electric and magnetic resonances overlap are determined by the geometrical parameters of the structure , as well as by the specific materials used .
therefore , it is possible to tune the wavelength of overlapping by appropriate choice of these parameters .
this gives the design a great degree of freedom , and make it plausible to operate at different frequencies . as an example ,
a ag@ge system with @xmath51 nm is presented ( figures [ fig2](e)-(h ) ) . in this case resonances overlap within 1500 nm and 1620 nm for @xmath55 nm .
the particluar case with @xmath52 nm is plotted in figure [ fig2](h ) . in figure [ fig3 ] , the near - field pattern at the combined magnetic ( shell ) and electric ( core ) resonance is shown for the ag@si system .
also plotted are the responses of the system to a purely electric and magnetic excitation .
these were obtained by placing a perfect mirror at a distance @xmath56 and @xmath2 , respectively , from the center of the structure .
the distinctive behaviour of both contributions is preserved in the combined electromagnetic resonance , revealed through the rotating field confined within the si shell ( as in figure [ fig1](e ) ) , together with the dipolar lspr resonance of the ag core .
note that , indeed , the electric displacement in the core and the shell rotates in opposite directions and that the magnetic near field pattern , figure [ fig3](b ) , can be explained as a combination of the electric and magnetic contributions .
near - field patterns for core - shell with ge covers are similar to those shown here .
nm and @xmath57 nm . ( a ) norm of the electric field in @xmath58 scale at combined electromagnetic resonance and when the system is excited by a purely electric excitation ( up - right , e ) and by a purely magnetic one ( down - right , h ) .
colour scale , incidence and polarization are preserved in all figures .
( b ) out - of - plane component of @xmath43 , together with the electric displacement ( @xmath59 ) field in black arrows at combined electromagnetic resonance and when the system is excited by a purely electric excitation ( up - right , e ) or a purely magnetic one ( down - right , h ) .
colour scale , incidence and polarization are preserved in all figures . ]
what do we expect for a material consisting of such core - shell nanostructures ?
in general it is highly non - trivial to extract the effective constitutive parameters of a metamaterial . here
, two different methods are applied to obtain the effective parameters of a metamaterial composed by the core - shell structures presented . in the first one
we will assume a random arrangement of the constituents , while in the second a metamaterial made of a cubic arrangement of them will be studied , both leading to a negative - index behaviour within certain wavelengths . for composites made of a cubic or random arrangement of dipolar particles ,
lorentz - lorenz theory is widely used @xcite , leading to the well known clausius - mossotti formulas relating the effective permittivity and permeability with the polarizabilities of the particles and the filling fraction @xmath60 , where @xmath61 is the number of particles per unit volume : @xmath62 where @xmath63 and @xmath64 are the electric and magnetic polarizabilities of the spherical particles . for the core - shell structures considered @xmath65 .
therefore , we expect them to be well within the approximation considered in the theory , thus behaving essentially as electric and magnetic point dipoles .
the electric and magnetic polarizabilities , @xmath66 and @xmath67 , respectively , are directly proportional to the scattering coefficients @xmath42 and @xmath34 [ factor @xmath68 . in figure
[ fig4](a ) and ( b ) , we have plotted the polarizabilities for a core - shell configuration of @xmath50 nm and @xmath53 nm .
it is clear from the graph that there is a spectral region where negative electric and magnetic polarizabilities are obtained .
interestingly , as mentioned before , no particular arrangement for the constituents is necessary to build the metamaterial , since the resonant electric and magnetic responses arise from each core - shell structure separately .
thus , assuming a random distribution we can compute the effective parameters from ( [ cm ] ) for different filling fractions . figures [ fig4](c)-(e ) depict the calculated effective permittivity , permeability , and refractive index for metamaterials made up by these ag@si core - shells with filling fractions @xmath69 , @xmath70 and @xmath71 .
the metamaterial has simultaneously negative permittivity and permeability for filling fractions higher than @xmath69 . for a filling fraction of @xmath71 , the system , moreover ,
has @xmath72 , although this relatively high filling fraction would be in the limit of validity of clausius - mossotti formulas . in order to quantify the losses of the system one can compute the so called figure of merit , defined as @xmath73 .
the computed @xmath74 in the left - handed spectral region for this filling fraction is depicted by the black curve , reaching a maximum value of @xmath75 , corresponding to @xmath76 . if ge is used instead of si , the nim behaviour starts with lower filling fractions , due to a stronger electric and magnetic responses ( see figure [ fig5 ] and compare the polarizabilities with those of figure [ fig4 ] ) .
superlensing capabilities are predicted for filling fractions lower than @xmath70 ( figures [ fig5](c)-(e ) ) .
again , the @xmath77 for the highest filling fraction is plotted as a black curve . in this case
it reaches a maximum value of @xmath78 , corresponding to @xmath79 . although being far from the best @xmath77 values reported for double - fishnet metamaterials ( @xmath80 @xcite , the predicted values are reasonably good , with the obvious advantage of isotropy of this proposal .
nm and @xmath53 nm and effective parameters for a metamaterial composed of a random arrangement of these structures .
( a ) real ( red ) and imaginary ( blue ) parts of the electric polarizability .
( b ) real ( red ) and imaginary ( blue ) parts of the magnetic polarizability . ( c)-(e ) real ( red ) and imaginary ( blue ) parts of the effective permittivity ( c ) , permeability ( d ) and refractive index ( e ) of a metamaterial with several filling fractions : @xmath69 ( dotted ) , @xmath70 ( dashed ) and @xmath71 ( continuous ) . the black curve and the inset in ( e ) represent the @xmath77 for the highest filling fraction in the left - handed spectral region . ]
nm and @xmath51 nm and effective parameters for a metamaterial composed of a random arrangement of these structures .
( a ) real ( red ) and imaginary ( blue ) parts of the electric polarizability .
( b ) real ( red ) and imaginary ( blue ) parts of the magnetic polarizability . ( c)-(e )
real ( red ) and imaginary ( blue ) parts of the effective permittivity ( c ) , permeability ( d ) and refractive index ( e ) of a metamaterial with several filling fractions : @xmath81 ( dotted ) , @xmath82 ( dashed ) and @xmath70 ( continuous ) . the black curve and the inset in ( e ) represent the @xmath77 for the highest filling fraction in the left - handed apectral region . ] to further test the metamaterial design , we consider now core - shell nanospheres arranged periodically in the vertices of a cubic lattice . we take germanium as the high permittivity coating and choose the geometrical parameters as @xmath51 nm and @xmath52 nm .
we consider the case in which the cubic lattice has a period @xmath83 nm ( corresponding to a filling fraction @xmath84 ) .
full numerical simulations ( see section methods ) of propagation across infinite slabs of different thicknesses ranging from 9 to 13 unit cells are carried out to ensure convergence .
bloch boundary conditions are applied in the interfaces with adjacent cells .
therefore , coupling between structures ( with an interparticle distance of only 5 nm ) is fully considered .
the spectral range of the simulation is 1.28 to 1.7 @xmath1 m and the wave impinges at normal incidence .
the effective material parameters are retrieved from the complex reflection and transmission coefficients through the usual equations found in the literature @xcite .
nm and @xmath52 nm arranged in the vertices of a cubic lattice with period @xmath83 nm , corresponding to a filling fraction @xmath84 .
( b ) detail of the normalized near electric field norm ( color scale , red@xmath851 , blue@xmath850 ) and electric displacement field ( indicated by white arrows ) in the periodic structure . ]
while the imaginary part is unambiguously computed from the simulation , once a sufficiently high number of unit cells in the propagation direction have been considered , the real part of the index of refraction must be determined with much more care .
concretely , one must pay special attention to ensure that the retrieved refractive index fulfills causality and passivity . in this way we have obtained two bounding values for the real part of the index ,
so it fulfills the two mentioned requirements . for a detailed description on how the index was obtained ,
see section methods . in figure
[ fig6 ] , the obtained bounds for the refractive index are plotted . with this retrieval technique , a negative index spectral region
is predicted . in the more conservative prediction , corresponding to the higher bound , the real part of the index reaches a minimum value of @xmath86 , with a @xmath87 , corresponding to a wavelength of 1.347 @xmath1 m . although the values for the index are lower than those predicted by clausius - mossotti , the negative index frequency band is present and confirms the core - shell design as meta - atom candidate for building a negative index metamaterial .
it needs to be pointed out that we are far from the limit of filling fractions that can be obtained with non - overlapping spheres ( cannonball pile ) . since for interparticle distances as low as 5 nm
, we found that coupling does not prevent the appearance of the magnetic resonance , we expect to reach higher negative values of the index in other periodic configurations with higher filling fractions .
further study of periodic configurations with other homogenization approaches is intended for future work .
obtaining a three - dimensional isotropic metamaterial having negative index of refraction in the optical part of the spectrum has been one of the major challenges for scientist and engineers devoted to electromagnetism in the last decade .
here we presented a new design based on core - shell nanospheres that operates in the near - infrared , which tackle many of the previously found problems , namely isotropy , polarization independence and lack of three - dimensionality . in our system , the effective response of the metamaterial is due to every isolated `` meta - atom '' .
therefore , no particular arrangement of the constituents is needed .
specifically , we demonstrated with realistic materials that , for a random arrangement , the system achieves double - negative index of refraction for different filling fractions and that super - resolution is possible .
we also tested the validity when a very simple periodic realization is assumed .
although the achieved values in the latter case are worse than predicted by clausis - mossotti for the same filling fractions , we havent explored here some other periodic configurations that are expected to give a stronger response . with regard to building these metamaterials ,
current silicon fabrication techniques allow the realization of complex nanostructures such as hollow nanospheres @xcite and opals @xcite . in some of the processes ,
the starting point are silica ( sio@xmath88 ) nanostructures , as is the case in @xcite , in which si opals are fabricated by magnesiothermic reduction of sio@xmath88 opals . since silver nanospheres
have been successfully covered with sio@xmath88 in variable thicknesses @xcite , there are plausible ways to realize the metamaterial proposed here , at least with si covers . concerning the fabrication of ge shells instead , a layer of a different material , suitable to grow it , can be added between the core and the shell .
this layer , if thin , would not affect excessively the physical response of the system , opening the possibility of fabrication of this system as well .
the underlying physical principles can of course be exploited at lower ( far - ir and terahertz ) frequencies , at which some dielectric materials exhibit very large refractive indices and certain materials ( e.g. polar crystals , doped semiconductors ) behave as plasmonic metals .
therefore , the results presented pave the way towards potential isotropic three - dimensional optical metamaterials designed on the basis of the physics underlying the doubly - resonant metallo - dielectric configuration .
extinction and scattering cross sections based on the extended mie theory were calculated with wolfram mathematica 8 .
the near - field plots in figures [ fig1 ] and [ fig3 ] were calculated using the rf module of comsol multiphysics v4.0a . the computational domain consisted of four concentric spheres which defined four subdomains .
the radii of the spheres were @xmath16 , @xmath17 , @xmath89 and @xmath90 . from inner to outer
, the subdomains represented the core , the shell and the embedding medium ( air ) .
the last domain was set to a spherical pml , which absorbed all scattered radiation .
the incident radiation was defined as a plane wave .
the mesh was constructed with the software built - in algorithm , which generates a free mesh consisting on tetrahedral elements .
the maximum element edge size was set to 15 nm in the whole core - shell structure , with a growth rate of 1.35 , meaning that elements adjacent to a given one should not be bigger than 1.35 times the size of it . for the pml and air domains ,
the maximum element edge size was 50 nm . all mesh sizes are below the value recommended in the program specifications , which sets a maximum edge size of @xmath91 of the effective wavelength for a correct meshing .
finally , for the simulation of wave propagation along a n - unit - cell thick slab , comsol was also used . in this case , the computational domain consisted of two concentric spheres representing the core - shells , and a right rectangular prism of width and height @xmath83 nm , and length @xmath92 .
bloch periodic boundary conditions were set in the directions perpendicular to the propagation and two ports activated in the direction of propagation that allow to compute the reflection and transmission complex coefficients . in these cases
the maximum element edge size was set to 40 nm in the core - shell subdomains , 50 nm in the ports boundaries , and 100 nm in the boundaries where periodic boundary conditions were applied .
a correct solution of the problem requires an identical meshing for the pairs of boundaries where periodic boundary conditions are applied . in all cases pardiso solver
was used . as an example , for 9-unit - cell slab simulation , the total mesh consisted of 149064 elements , and the calculation involved 948768 degrees of freedom , requiring almost 18 gb of memory .
the effective refractive index for the metamaterial structure , shown in figure [ fig6 ] , were calculated from complex transmission and reflection coefficients following @xcite . as mention before , we performed simulations of wave propagation across infinite slabs of different thicknesses .
after a sufficient number of unit cells is considered we get convergence of the results .
we needed to consider up to 9 unit cells in the direction of propagation to get convergence .
then , we performed the simulation for 11 and 13 unit - cell - thick slabs .
it is well known that , when extracting the effective index through transmission and reflection coefficients , an ambiguity arises related to the branches of the complex logarithm function .
this ambiguity only affects the determination of the real part of the index , the imaginary one being univocally defined .
it is a common assumption that the real part of the index can be found plotting different branches for different thicknesses .
the physical index is assumed to be the one independent of the thickness of the slab . however , this assumption may be ambiguous as well .
the reason is that , if one plots a sufficiently high number of branches one will find more than one unique index independent of the thickness . in figure [ fig7 ]
this situation is shown .
we have plotted a very large number of branches corresponding to the ag@ge core - shell periodic configuration of section 3.2 .
it can be clearly seen that there is not only one case in which the index is independent of the thickness .
nm and @xmath52 nm arranged in the vertices of a cubic lattice with period @xmath83 nm , corresponding to a filling fraction @xmath84 .
the branches that are independent of the thickness of the slab are indicated by a surrounding ellipse .
the dashed blue line represents the real part obtained from the imaginary part by kramers - kronig relations , plus a constant factor to fit the only thickness - independent index showing good agreement . ] therefore , although this requirement of thickness - independence of the retrieved index is clearly physically acceptable , we should not take it as a sufficient condition .
in fact , we should impose more conditions to the retrieved index .
it must , no doubt , fulfill the basic requirement of causality , expressed mathematically by the kramers - kronig relations . since , fortunately , we can unambiguously determine the imaginary part of the index , it is possible to apply kramers - kronig relations to get the real part of the index @xcite .
we need to point out that , actually , by this technique one can only compute the corresponding real part up to a constant factor . the value of this factor , interpreted physically as the value of the index when the frequency tends to infinity , needs to be determined somehow else .
the question is , then , if any of the indices that do not depend on thickness also fullfils causality . in figure
[ fig7 ] , the real part of the index , retrieved through kramers - kronig relations applied to the imaginary part , is plotted on top of the only thickness - independent index for which the fitting is in good agreement ( the constant factor added is 6.35 ) .
it is apparent that the agreement is quite good .
it should be noticed that this index is the only one for which the fitting is good , thus fulfilling the the causality constrain .
however , one more constrain must be imposed in order to get an acceptable effective index .
the resulting effective medium should be passive .
this condition demands no spontaneous generation of energy , and mathematically relates the real and imaginary parts of the effective index , with the real and imaginary parts of the effective impedance of the slab .
this impedance can be computed also from the complex reflection and transmission coefficients @xcite . applying this condition
we find that none of the indices retrieved directly from the transmission and reflection complex coefficients fulfill both causality and passivity .
nevertheless , one can still compute a physically acceptable effective index of refraction . if one uses the passivity condition , it is possible to impose upper and lower bounds to the constant factor in kramers - kronig relations . an index with a real part given directly by kramers - kronig relations , applied to the univocally computed imaginary part , and a constant factor between the two bounds
will necessarily fulfill both conditions .
we found that , in fact , the upper and the lower bounds imposed by passivity are quite close in value , being a restrictive condition on the possible values of the index of refraction .
the two boundary values for the index are plotted in figure [ fig6](a ) .
the authors acknowledge support both from the spain ministerio de ciencia e innovacin through the consolider - ingenio project emet ( csd2008 - 00066 ) and nanoplas ( fis2009 - 11264 ) , and from the comunidad de madrid ( grant microseres p2009/tic-1476 ) .
r. paniagua - domnguez acknowledges support from csic through a jae - pre grant .
we are indebted to an anonymous referee for helpful suggestions on the convergence of the comsol results in figure 6 .
10 smith , d. r. , padilla , w. j. , vier , d. c. , nemat - nasser , s. c. and schultz , s. _ phys .
_ * 2000 * , 84 , 4184 - 4187 .
shalaev , v. m. _ nature photon . _ * 2007 * , 1 , 41 - 48 .
soukoulis , c.m . and
wegener , m. _ nature photon . _ * 2011 * , 154 , 1 - 7 .
zhou , j. , koschny , th . ,
kafesaki , m. , economou , e. n. , pendry , j. b. and soukoulis , c. m. _ phys .
lett . _ * 2005 * , 95 , 223902 .
al , a. , salandrino , a. , and engheta , n. _ opt .
, * 2006 * , 14 , 1557 .
gney , d. . , koschny , t. and soukoulis , c.m .
b _ * 2011 * , 83 , 045107 . dolling , g. , wegener , m.,soukoulis , c.m . and linden , s. _ opt . lett . _ * 2007 * , 32 , 53 - 55 .
dionne , j. a. , verhagen , e. , polman , a. and atwater , h. a. _ opt .
express _ * 2008 * , 16 , 19001 - 19017 .
burgos , s. p. , de waele , r. , polman , a. , atwater , h. a. _ nature materials _ , * 2010 * , 9 , 407 - 412 .
lezec , h. j. , dionne , j. a. and atwater , h. a. _ science _ * 2007 * , 316 , 430 - 432 .
wheeler , m. s. , aitchison , j. s. , and mojahedi , m. _ phys .
b _ * 2005 * , 72 , 193103 .
quian zhao , lei kang , du , b. , zhao , h. , xie , q. , huang , x. , li , b. , zhou , j. and li , l. _ phys .
lett . _ * 2008 * , 101 , 027402 .
vynck , k. felbacq , d. centeno , e. cabuz , a. i. cassagne , d. and guizal , b. _ phys .
lett . _ * 2009 * , 102 , 133901 .
yannopapas , v. and moroz , a. _ j. phys . : condens
. matter _ * 2005 * , 17 , 37173734 .
yannopapas , v. and vitanov , n.v .
b _ * 2006 * , 74 , 193304 .
yannopapas , v. _ phys .
b _ * 2007 * , 75 , 035112 .
kussow , a - g . ,
akyurtlu , a. and angkawisittpan , n. _ phys .
b _ * 2008 * , 245 , 992 . seo , b - j . ,
ueda , t. , itoh , t. and fetterman , h. _ app .
_ * 2006 * , 88 , 161122 .
rockstuhl , c. , menzel , c. , mhlig , s. , petschulat , j. , helgert , c. , etrich , c. , chipouline , a. , pertsch , t. and lederer , f. _ phys .
b _ * 2011 * , 83 , 245119 .
smith , d. r. , vier , d. c. , koschny , t. and soukoulis , c. m. _ phys .
e _ * 2005 * , 71 , 036617 .
chen , x. , grzegorczyk , t. m. , wu , b. , pacheco , j. jr . and
kong j. a. _ phys .
* 2004 * , 70 , 016608 .
aden , a. l. and kerker , m. j. appl .
* 1951 * , 22 , 1242 .
garca - etxarri , a. , gmez - medina , r. , froufe - prez , l. s. , lpez , c. , chantada , l. , scheffold , f. , aizpurua , j. , nieto - vesperinas , m. , and senz , j. j. _ opt .
express _ * 2011 * , 19 , 4815 .
comsol multiphysics v4.0a .
jelinek , l. , and marqus , r. _ j. phys . : condens
. matter _ * 2010 * , 22 , 025902 .
bohren , c. f. and huffman , d. r. _ absorption and scattering of light by small particles _
( wiley - interscience publication , new york , 1983 ) .
johnson , p. b. and christy , r. w. _ phys .
b _ * 1972 * , 6 , 4370 .
wheeler , m. s. , aitchison , j. s. , chen , j. i. l. , ozin , g. a. , and mojahedi , m. _ phys .
b _ * 2009 * , 79 , 073103 .
garcia - meca , c. , hurtado , j. , mart , j. , martnez , a. , dickson , w. and zayats , a. v. _ phys .
lett . _ * 2011 * , 106 , 067402 .
yao , y. , mcdowell , m. t. , ryu , i. , wu , h. , liu , n. , hu , l. , nix , w. d. and yi cu .
_ nano lett . _ * 2011 * article asap .
ibisate , m. , golmayo , d. , and lpez , c. _ avd . mat . _ * 2009 * , 21 , 2899 .
aslan k. , wu m. , lakowicz j. r. and geddes c. d. _ j. am .
soc . _ * 2007 * , 129 , 1524 - 1525 .
ohta , k. and ishida , h. _ appl .
spectrosc . _ * 1988 * , 42 , 952 - 957 . | materials showing electromagnetic properties that are not attainable in naturally occurring media , the so called metamaterials , have been lately , and still are , among the most active fields in optical and materials physics and engineering . among those properties ,
one of the most attractive is the sub - diffraction resolving capability predicted for media having index of refraction of -1 .
here we propose a fully 3d , isotropic metamaterial with strong electric and magnetic responses in the optical regime , based on spherical metallo - dielectric core - shell nanospheres .
the magnetic response stems from the lowest , magnetic - dipole resonance of the dielectric shell with high refractive index , and can be tuned to coincide with the plasmon resonance of the metal core , responsible for the electric response .
since the response does not originate from coupling between structures , no particular periodic arrangement needs to be imposed .
moreover , due to the geometry of the constituents , the metamaterial is intrinsically isotropic and polarization independent .
it could be realized with current fabrication techniques with materials such as silver ( core ) and silicon or germanium ( shell ) . for these particular realistic designs , the metamaterials present negative index in the range within @xmath0 @xmath1 m . |
high - mass star formation is poorly understood because theory has yet to resolve the balance of radiation pressure , gravity , magnetic energy , and thermal heating ( e.g. , mckee & tan 2002 ) .
moreover , examples amenable to detailed study but young enough to have not yet formed hii regions are rare . first , massive ysos are deeply embedded , form in clusters , and heat large amounts of surrounding dust and gas . as a result ,
the signatures of disks and outflows at small radii are easily lost in `` background clutter . ''
second , regions of high - mass star formation are sufficiently distant ( @xmath1pc ) that infrared and radio observations of thermal gas and dust are often confusion - limited .
third , massive young stellar objects evolve rapidly , and by the time the surrounding medium is dispersed , accretion disk and outflow structures have been at least partly dispersed . in two of the best understood cases , [email protected] ( shepherd & kurtz 1999 ) and iras20126 + 4104 ( zhang et al . 1998 ; cesaroni et al .
1999 ) , disk - like structures have been detected , but the rotation curves are barely resolved and the emission arises at large enough radii ( @xmath3au ) that in depth study is difficult .
the infrared kleinemann - low ( kl ) nebula in orion is the nearest ( @xmath4pc ) , and most heavily studied region of high - mass star formation .
the kl region is crowded , containing 16 identified mid - infrared peaks across @xmath5au ( gezari , backman , & werner 1998 ) .
offset @xmath6 south of the prominent peak , irc2 , is radio source i , first detected by churchwell et al .
it has no infrared counterpart , but it does power compact distributions of sio and h@xmath0o masers , and it has been identified as a probable deeply embedded , massive protostar or yso ( gezari 1992 ; menten & reid 1995 ) .
the strongest sio maser emission arises from the v=1 vibrational state . because of excitation requirements ( @xmath7k , @xmath8@xmath9 ) , it must originate close to the yso .
the first vlbi maps revealed an x - shaped sio emission locus extending outward 2070au from source i. the x also lay at the center of a @xmath10au expanding patch of h@xmath0o masers distributed in two lobes each comprising red and blueshifted emission and bracketing source i ( greenhill et al .
1998 ; see also gaume et al .
1998 ; doeleman , lonsdale , & pelkey 1999 ) . taken alone
, the distribution of sio emission could have been interpretted equivalently in two ways .
first , it could represent the turbulent limb of a high - velocity biconical outflow with a southeast - northwest axis .
second , it could represent hot material close to the top and bottom surfaces of an edge - on disk with a northeast - southwest rotation axis .
greenhill et al . and doeleman et al . adopted the biconical outflow model , in part because the high - velocity outflow observed in the kl region on scales of @xmath11au ( e.g. , allen & burton 1993 ; schultz et al . 1999 ) subtended the opening angle of the putative sio maser cone .
in addition , greenhill et al . ( 1998 ) noted that the distributions of proper motions and line - of - sight velocities among the surrounding h@xmath0o masers were not readily consistent with the edge - on disk model and proposed instead that the h@xmath0o masers lay on the surface of a nearly edge - on , inflating , equatorial doughnut - like shell . this shell could arise from a slow stellar equatorial wind advancing into the surrounding medium or from photo - evaporation of an accretion disk .
because the major axis of this shell was aligned with the the so - called `` 18 kms@xmath12 '' outflow in the kl region ( e.g. , genzel & stutzki 1989 ) , it appeared that source i might drive the two most prominent outflows in the region . despite circumstantial evidence tying source i to large scale outflows in the kl region
, ambiguity has remained .
on the one hand , menten & reid ( 1995 ) noted that another infrared source , n , lies closer to the previously estimated center of the flows ( genzel et al .
1981 ) . moreover , high - velocity gas far to the northwest is blueshifted , while the corresponding outflow cone near source i is redshifted . on the other hand , the compact bipolar radio lobes of source n are oriented north - south , @xmath13 from the principal axes of the low and high - velocity large scale flows in the kl region . to better understand the kl region and to test the biconical outflow model proposed for source i , we remapped the sio and h@xmath0o maser emission , covering three @xmath14 transitions of sio ( v=0 , 1 , & 2 ) and the full velocity range of h@xmath0o emission .
( the original maps presented by greenhill et al . were significantly limited in dynamic range , especially at velocities close to the systemic velocity . )
we measured the proper motion of the sio v=1 & 2 emission , observing monthly over @xmath15 years , which is @xmath16 of the dynamical crossing time of the outflow .
here we discuss proper motions over a four month interval .
we observed the v=1 & 2 @xmath17 transition of sio with the vlba , and the v=0 @xmath17 transition of sio and @xmath18 transition of h@xmath0o with the vla in its largest configuration and including the nearby pie town vlba antenna .
we obtained angular resolutions of @xmath19milliarcseconds ( mas ) with the vlba and @xmath20mas with the vla .
the spectral channel spacings in the vlba and vla imaging were @xmath21kms@xmath12 and @xmath22kms@xmath12 , respectively . using the vlba , we observed the sio v=1 & 2 emission ( from @xmath23 to + 25kms@xmath12 ) simultaneously to enable registration of both lines to @xmath24mas . because emission from source i itself is thermal , it could not be detected with the vlba
. to locate the position of the yso on the vlba maps , we convolved the maps with a circular beam comparable in size to the beam of the vla at @xmath25 mm and overlayed the degraded vlba map on the vla map of menten & reid ( 1995 ) , who detected and registered the sio v=1 and thermal continuum emission of the yso .
the uncertainty in this registration is @xmath26mas . using our vla observations
, we measured the positions of the sio v=0 masers ( @xmath27 to + 22kms@xmath12 ) with respect to the v=1 emission , which was observed simultaneously so as to provide an astrometric reference .
the uncertainty in position relative to source i is @xmath26mas .
the positions of the h@xmath0o masers ( @xmath27 to + 16kms@xmath12 ) were measured with respect to a nearby quasar and compared to the absolute position of source i ( menten & reid 1995 ) .
the uncertainty in this comparison is 30mas .
our more extensive mapping of the sio and h@xmath0o maser emission now supports the edge - on disk model previously discarded .
we propose that the v=1 & 2 sio maser emission traces material streaming in a rotating funnel - like flow from the upper and lower surfaces of an accretion disk of a massive yso .
the v=0 sio and h@xmath0o masers lie in a bipolar outflow along the disk rotation axis . in the most recent observations ,
the v=1 & 2 sio emission traces an x as before ( figure1 ) , but we note two important new details .
first , there is a `` bridge '' of maser emission extending from the base of the south arm to the base of the west arm , and there is a gradient in line - of - sight velocity along the bridge .
second , the arms are not radial
. these new findings may be consequences of our having now mapped the v=2 as well as v=1 emission .
( for instance , the bridge is outlined principally by v=2 emission . )
both findings are difficult to explain in the context of the biconical flow model .
however , they are readily explained in the context of the edge - on disk model .
the bridge and associated velocity gradient are natural signatures of emission from the front side of a rotating disk that is tipped down slightly to the southwest .
the canting of the arms , so that they are not radial , is also suggestive of reflection symmetry about a plane ( i.e. , disk ) . in order to estimate the proper motions of maser clumps
, we identified one in each quadrant of the x that persisted in each epoch and maintained the same line - of - sight velocity .
we overlayed images for the four epochs and registered them to achieve zero mean proper motion for the source as a whole .
most proper motions are 10 to 15kms@xmath12 , though some clumps are nearly stationary on the sky .
( one kms@xmath12 over four months corresponds to @xmath29 vlba beamwidth . )
the maximum observed 3-d velocity is @xmath30kms@xmath12 .
the proper motions of maser clumps are chiefly along the four arms of the x ( figure1 ) . in the bridge ,
the motions are both outward and tangential , indicative of rotation and strongly in support of the edge - on disk model ( chandler et al . , in prep ) .
we suggest that the arms of sio emission represent the limbs of a bipolar funnel - like outflow ( figure2 ) , probably the turbulent shocked interface between outflowing and accreting material . in a rotating system , strong maser emission is expected along the limbs because that is where the longest maser gain paths lie owing to projection effects .
additional emission from the bridge may mark the nearside of the outflow wall , where at the base , higher densities or temperatures could compensate for otherwise shorter gain paths .
the dynamical mass of the yso is difficult to estimate because the maser material is not in simple keplerian rotation .
if the observed 3-d velocity is greater than the escape velocity , then the enclosed mass is @xmath31m@xmath32 , for a radius of 25au and velocity of 25kms@xmath12 .
we note that the 3-d motion of the maser material is helical
. this may indicate the presence of a strong magnetic field that is probably anchored to the accretion disk , since massive stars are radiative and may not generate their own field .
if the magnetic and kinetic energy densities are of the same order , then the field is on the order of 1 g , for a gas density of @xmath33@xmath9 . the v=0 sio and h@xmath0o maser emission arises chiefly from two `` polar caps '' that subtend the @xmath35 opening angle of the outflow along the disk rotation axis ( figure2 ) .
the velocity structure of the emission is somewhat disorganized and difficult to model in detail .
however , the two lobes ( northeast and southwest ) display the same ranges of velocity , indicative of outflow in the plane of the sky . within each lobe , the bulk of the emission displays a red - blue asymmetry across the rotation axis in the same sense as the disk rotation close to source i , although further study is required
. otherwise , it may signify coupling between the velocity field downstream in the outflow and the underlying accretion disk , possibly via magnetic processes .
substantial overlap of h@xmath0o maser and sio v=0 emission along the line of sight is surprising because maser action in each species requires quite different densities , @xmath36 @xmath9 vs @xmath37 @xmath9 , respectively
. moreover , because molecule reformation on dust grains is essential for h@xmath0o maser action behind shocks , gas phase sio would be depleted . in principle
, the h@xmath0o and v=0 sio maser volumes could be intermingled if the flow were inhomogeneous and if shock induced grain sputtering enhanced gas phase concentrations of sio .
however , the requisite @xmath38 density contrast between emitting regions would be difficult to explain in the apparent absence of high - velocity ( @xmath39kms@xmath12 ) gas motions .
we suggest that the wall of the outflow traced by v=1 & 2 sio masers at radii @xmath40au extends to larger radii where it supports h@xmath0o maser emission .
the outflow itself has low enough density to support the v=0 sio maser emission , which lies along the same line of sight as the h@xmath0o emission in projection .
the ordered velocity structure of the v=1 & 2 sio maser emission and the relatively complicated velocity structure of the h@xmath0o and v=0 sio maser emission at larger radii may be a consequence of the increased effects turbulence or collisions with inflowing or ambient material downstream in the outflow .
we have fully resolved the structure and dynamics of material at radii of 20500au from radio source i in the orion kl region .
our maps provide the most detailed picture yet of molecular material so close to a massive yso .
we propose a new model for source i in which the accretion disk is edge - on and a wide - angle wind feeds a funnel - like rotating bipolar outflow .
consequently , the case of source i provides strong evidence that accretion in high - mass star formation proceeds via orderly disk - mediated accretion as opposed to coalescence of low mass stars ( bonnell et al . | high - mass star formation is not well understood chiefly because examples are deeply embedded , relatively distant , and crowded with sources of emission . using vla and vlba observations of h@xmath0o and sio maser emission ,
we have mapped in detail the structure and proper motion of material 20500au from the closest high - mass yso , radio source i in the orion kl region .
we observe streams of material driven in a rotating , wide angle , bipolar wind from the surface of an edge - on accretion disk .
the example of source i provides strong evidence that high - mass star formation proceeds via accretion . 1.25 in .125 in .25 in |
when a flat thin sheet of material is subjected to increasing pressure , it eventually buckles , crumples , or cracks . by buckling , the sheet releases stress over wavelengths much larger than its thickness . during the past years
there has been increasing interest in phenomena , such as crumpling , where the deformation goes from an initial uniform state to a localized region occupying an arbitrarily small fraction of the sample @xcite .
monolayers of surface - active molecules ( surfactants ) , adsorbed on a liquid , are found in many systems containing water
air or water oil interfaces @xcite .
such monolayers exhibit a rich variety of collapsed structures under lateral compression , most of which occur on an intermediate scale between the macroscopic one and the molecular thickness of the layer .
fluid monolayers collapse into disks , tubes , or pearls - on - string structures , depending on spontaneous curvature and charge of the lipid monolayer @xcite . like elastic sheets , many solid - like monolayers crack @xcite or buckle @xcite under pressure , yet other solid - like monolayers fail by abrupt buckling into straight , micron - wide folds @xcite ( see fig .
[ fig_buckle ] for an example ) .
in addition , liquid - like monolayers may form micron - scale vesicular objects of various shapes @xcite or giant convoluted folds @xcite .
this type of folding is believed to be driven by the interfacial energy gained from the contact across the two sides of the fold @xcite .
thus , the failure of surfactant monolayers under lateral pressure displays distinctive mechanical behaviors , which crucially depend on the in - plane rigidity @xcite and viscoelasticity @xcite .
lipid monolayers and bilayers are the material of choice for spatial partitioning in living matter , such as cells and compartments within them .
these partitions are often observed to fold and wrinkle under stress . in particular , lipid monolayers that model the expanding and contracting sacs in an animal s lung exhibit the abrupt folding signature described in refs .
@xcite .
this abrupt buckling is particularly puzzling . the abrupt relaxation motions or
jerks " have been extensively characterized for a particular system @xcite , but their distinctive length and time scales have not been explained . why do the jerks occur over a time scale of about 0.1 seconds far from any molecular or apparatus time - scales ?
why does the motion stop suddenly , before the driving stress is relaxed ? why do the jerks show a robust characteristic displacement of a few microns ?
an intriguing hypothesis to explain this micron length scale was recently proposed by pocivavsek et al @xcite .
these authors recalled that any thin sheet under compression on a liquid substrate buckles at a well - defined wavelength @xmath0 . for lipid monolayers
the expected value of @xmath0 lies in the micron range .
thus it is of the same order as the characteristic displacement of the jerk relaxations .
these authors also noted that the incipient wrinkles at wavelength @xmath0 are unstable against folding , in which the excess wrinkled material from throughout the sample is concentrated into a single loop or fold .
this paper aims to account for the dynamical features of the jerk motion . in section [ sec_system ]
we argue that the monolayer may be viewed as a broad , thin slab that translates almost rigidly during the jerk motion .
then in section [ sec_dynamics ] we describe how such a slab should respond to horizontal forcing , accounting for progressive viscous entrainment of the fluid subphase . using this result ,
combined with experimental observations , we set a lower bound on the kinetic energy of the jerks .
we then survey the possible forces that might give rise to this kinetic energy , notably the energy released by folding in the mechanism of ref .
even upper limits of this folding energy are far less than this observed kinetic energy . in section [ sec_discuss ]
we discuss effects that might account for this kinetic energy .
for definiteness we focus our study on the jerks analyzed in ref .
@xcite and re - examined in ref .
we begin by reviewing the parameters of this system .
the monolayer was a 7:3 mixture of dipalmitoylphosphatidylcholine ( dppc ) and dioleoylphosphatidylglycerol ( popg ) spread on pure water at room temperature in a @xmath1-cm - long langmuir trough and viewed in a microscope in a 100-micron - wide field of view at conventional video frame rates .
it is compressed at 0.1 mm / sec to a nominal pressure of about 70 mn / meter before viewing .
the measured pressures are consistent with the slight overpressure needed for folding @xcite . in these conditions
this material microphase separates into a biphasic foam - like pattern of compact patches separated by narrow strips of different composition , visualized by a dilute fluorescent additive .
the patches are 15 - 25 microns in size .
the monolayer behaves mechanically as a solid , not a fluid ; the jerks move in the direction of the langmuir barrier motion .
rheological measurements in similar systems showed stress relaxation times of order @xmath2@xmath3 sec.@xcite .
the jerks vary statistically in their net displacement @xmath4 and their duration @xmath5 .
no statistical correlations between jerks were observed .
the displacements @xmath4 vary from a minimum of about 1.2 microns to several times larger , with an average of 2 microns @xcite .
the durations @xmath5 vary from a minimum of about @xmath6 sec to a few times longer , where the average is 0.12 sec .
it has been suggested @xcite that the larger jerks are cascades of elementary jerks , with this average @xmath4 and @xmath5 . in the estimates below
, we use these average @xmath4 and @xmath5 values .
we may simplify our description of the monolayer using three further features .
first , the monolayer may be assumed to translate _ rigidly _ outside the folding region over distances of several cm .
it thus entrains substrate fluid over these distances . in principle
compressibility could invalidate this assumption .
compressibility implies that an initial imbalance of membrane stress produces a compressional wave whose speed is given by , @xmath7 , where @xmath8 is the two - dimensional uniaxial compression modulus and @xmath9 is the mass density per unit area .
if forces are applied on a timescale @xmath5 , the resulting compression or expansion is confined to distances @xmath10 .
conversely , the compression is negligible and the body moves rigidly if its size @xmath11 is smaller than about @xmath12 .
thus to show that our monolayer translates rigidly , we must establish that @xmath13 is sufficiently large . for our case
the density @xmath9 is the density of the material to be accelerated in propagating the wave .
since the monolayer entrains substrate fluid as it moves , the density must take account of this subphase .
we may find a lower limit on this @xmath13 by using a lower limit to the modulus @xmath8 and an upper limit for the surface density @xmath9 .
accordingly we estimate @xmath8 by neglecting any compressional effects of the subphase
. measurements of the monolayer compression modulus range from 0.1 n / m to several n / m @xcite .
we use the conservative estimate of @xmath14
n / m below . as for the surface density @xmath9
we find an upper limit by including all the water that might be entrained .
as discussed in section [ sec_dynamics ] , an upper limit on the entrained density @xmath9 is given by @xmath15 , where @xmath16 ( @xmath17 kg / m@xmath18 ) is the density of the water subphase , and @xmath19 ( @xmath20 m@xmath21/sec ) is its kinematic viscosity .
thus in the jerk time of 0.12 sec , a local compression can propagate a distance @xmath22 m , i.e. even by our minimal assumption , information from the folding region has reached about half the sample ( 5 cm ) within the jerk time .
thus we expect only minor effects from compressibility of the sheet . in the calculations to follow we will assume @xmath23 cm is the size of the moving sheet .
a second feature of the monolayer is that its compressibility has negligible effect on the energetics of folding . as explained below
, the over - pressure @xmath24 released in the wrinkle - to - fold transition @xcite in this system must be smaller than @xmath25 mn / m .
given the large lower - bound modulus @xmath8 above , the compressive displacement @xmath26 in a system of length @xmath11 is smaller than @xmath27 micron .
it is thus much smaller than the observed displacements @xmath4 .
likewise , the compressive energy released is a small fraction ( @xmath28 ) of the folding energy .
a third feature of these monolayers is that the longitudinal propagation of the fold tips is much faster than the transverse folding and its accompanying jerk . within the temporal resolution of ref .
@xcite ( @xmath29 s ) the fold traverses the field of view ( @xmath30 @xmath31 m ) instantaneously ; this sets a lower bound of @xmath32 mm / s for its speed . in the system of ref.@xcite two types of longitudinal folds
were observed , the slower of which propagated at @xmath33 mm / s . comparing these values with the characteristic translation velocity , @xmath2
@xmath31 m/s @xcite , we see that the requirement is safely fulfilled .
these features allow us to consider a simplified two - dimensional problem of a thin elastic sheet moving over a semi - infinite viscous liquid ( fig .
[ fig_scheme ] ) .
current understanding @xcite attributes the jerks to the mechanical buckling or wrinkling instability of any elastic sheet that is floating on a liquid and is under compression @xcite . above a threshold pressure @xmath34 ,
the sheet distorts , at a cost of bending energy .
the wrinkles also produce a net upward and downward displacement of the liquid , thus increasing its gravitational energy . taking account of these costs ,
one finds a threshold pressure given by @xmath35 , where @xmath36 is the bending stiffness and @xmath37 is the acceleration of gravity .
the predicted buckling wavelength @xmath0 is given by @xmath38^{1/4}$ ] .
this buckling is unstable , leading to a release of the overpressure @xmath34 over a displacement @xmath39 comparable to @xmath0 .
though @xmath34 is too small to measure directly in our system , it can be estimated either using typical values of @xmath36 or by inferring @xmath36 from the observed jerk displacement . bending stiffness for lipid monolayers like our system
lie in the range @xmath40 j. @xcite .
this @xmath36 value implies an upper bound for @xmath34 : @xmath41 mn / m .
if instead we infer @xmath34 from the observed jerk displacements via @xmath42 , we obtain values of @xmath43 mn / m , i.e. 30 times smaller than our upper bound @xcite .
in what follows we will use the more conservative upper - bound " value of @xmath34 , namely @xmath44 mn / meter .
there are many forces in the system that exceed this value - the total pressure from the trough , drift flows , flows from faraway jerks , or residual flows from previous jerks .
however , these larger pressures lack the central feature needed to explain jerks .
the jerks clearly result from an instability , in which a small displacement @xmath4 leads to an increasingly _ unbalanced _ pressure . to summarize the above discussion , the sheet is assumed to deform in the @xmath45@xmath46 plane while remaining uniform along the @xmath47 axis .
it is laterally compressed by a two - dimensional , uniform pressure @xmath24 .
( in the experimental system @xmath24 is given by the actual pressure exerted at the boundaries minus the surface tension of the liquid . )
prior to instability the monolayer remains flat ( @xmath48 ) and responds to compression by slightly decreasing its actual length .
when the pressure exceeds @xmath34 , the monolayer wrinkles or folds out of the @xmath45@xmath47 plane ( @xmath49 ) . since the pressure @xmath34 produces negligible elastic compression , the total length @xmath11 is fixed , and the displacement @xmath4 along the @xmath45 axis is fully accounted for by the wrinkles or folds . the underlying liquid , having viscosity @xmath50 and mass density @xmath16 , occupies the region @xmath51 .
as noted above , the motion of the sheet depends strongly on the viscous entrainment of the fluid beneath it . in this section
we determine the motion taking account of this entrainment .
we begin with a simplified situation .
suppose we control the surface pressure and increase it above the instability pressure , @xmath52 .
we further suppose that as the sheet begins to fold , the buckled region is decreasing the pressure by a constant amount @xmath24 . at this moment
the forces on the sheet are unbalanced and it begins to accelerate . in all but the small folding region the sheet is horizontal and
its motion is a pure horizontal translation .
the flat sheet adjacent to the fold will translate laterally , thus creating a velocity profile @xmath53 in the underlying liquid . on larger scales this flow produces circulation of fluid in the sample .
we ignore such large scale gestures , since these only increase the kinetic energy and the discrepancy we are discussing .
the resulting viscous drag decreases with time , as the velocity profile penetrates deeper into the liquid and its gradient becomes less sharp .
we shall assume that the sheet s velocity is always equal to that of the water immediately beneath it .
that is , the fluid obeys a no - slip boundary condition .
we examine such assumptions in section [ sec_discuss ] below .
we may neglect the mass of the sheet relative to the much greater mass of the entrained fluid .
thus the problem is simplified to that of finding the velocity of a half infinite fluid due to a rigid surface moving with a constant lateral pressure .
the equation of motion for the fluid is : @xmath54 where @xmath19 is the kinematic viscosity given by the ratio of viscosity and density , @xmath55 , and a dot denotes derivative with respect to time .
the constant stress at @xmath56 implies a boundary condition of the form @xmath57 accompanying it are the condition of vanishing flow far away from the interface , @xmath58 and initial conditions of stationarity , @xmath59 eqs ( [ fluid_motion ] ) and ( [ pressure ] ) may be simplified by defining the laplace transform @xmath60 by @xmath61 : @xmath62 the solution in laplace space is @xmath63 and in the time domain , @xmath64 , \ \
\alpha\equiv z/(2\sqrt{\nu t } ) \label{solution_real}\ ] ] the displacement of the sheet is given by integrating the velocity on the surface , @xmath65 , @xmath66 since the accelerated mass is constantly increasing in time , the acceleration @xmath67 decreases with time as @xmath68 .
this leads to the unusual `` jerky '' increase of the velocity @xmath69 . we note that this predicted motion arises from basic hydrodynamics
; it does not depend on the source of the pressure @xmath24 .
we note also that a fixed fraction @xmath70 of the input power @xmath71 goes into kinetic energy ; the remainder goes into viscous dissipation . buckling relaxes stress at the tip of the fold
; we would therefore expect the driving pressure to depend on the displacement .
the equation for @xmath72 can readily be generalized to include such a @xmath73 .
we discuss it here for completeness .
@xmath74 where the first term on the right is the driving stress , the last term is the stress from the fluid , and @xmath75 is the two - dimensional mass density of the sheet .
for the fluid , eq .
( [ fluid_motion ] ) still applies , and the two are supplemented by a no - slip boundary condition at the interface , @xmath76 a condition of vanishing flow far from the surface ( eq . [ bc2 ] ) , and initial conditions of stationarity ( eq . [ bc3 ] ) .
we proceed by eliminating @xmath77 to obtain an equation for the sheet alone .
the velocity green function of the liquid is given by the boundary condition @xmath78 and eqs .
( [ bc2],[bc3 ] ) .
explicitly , @xmath79 the general solution with the no - slip boundary condition of eq .
( [ bc1 ] ) is obtained simply by integration , @xmath80 . inserting this expression in eq .
( [ motion1 ] ) and integrating by parts , produces an equation for the sheet alone : @xmath81 as before , we may neglect the inertial term on the left hand side . for a constant pressure , eq .
( [ power32 ] ) gives the same displacement as eq.([solution_real ] ) .
we may now compare the observed jerk motion with the motion expected from the folding forces .
it is straightforward to solve eq .
( [ power32 ] ) using the predicted ( quadratic ) @xmath73 @xcite .
the resulting motion proves to be much slower than the observed jerks .
we may quantify this discrepancy in two ways .
first we ask what constant pressure @xmath24 would be required to give the observed displacements @xmath4 in the observed time @xmath5 .
then we determine the observed kinetic energy and compare it to the energy available from folding . to find the pressure required to produce the observed jerks , we substitute @xmath82 microns and @xmath83 sec into eq .
( [ power32 ] ) , to obtain @xmath84 mn / m .
this is some 50 times larger than the upper - bound folding pressure @xmath85 mn / m obtained above .
the same discrepancy emerges if we compare the observed kinetic energy @xmath86 with the work @xmath87 done by the pressure difference @xmath73 .
both are proportional to the width @xmath88 of the jerking region .
the pressure @xmath73 is always smaller than @xmath34 throughout the folding .
thus @xmath89 n . other things being equal ,
the kinetic energy @xmath86 for a given average velocity @xmath90 is larger if @xmath5 is larger ( since larger @xmath5 implies a greater entrained mass ) .
thus we may obtain a lower bound on the kinetic energy by limiting @xmath5 to the observed duration of @xmath91 sec . for the moment we simply use the constant - pressure solution of eq .
( [ solution_real ] ) and use @xmath92 \label{max_kinetic}\ ] ] we examine this estimate in sec .
[ sec_discuss ] . for @xmath93
m , @xmath82 microns and @xmath83 sec , this gives @xmath94 n some 32 times our upper bound of the work @xmath87 supplied by folding .
thus even when one ignores the work that must go into viscous dissipation , the observed energy is far larger than what the folding energy , @xmath87 , can supply .
( taking account of the dissipated energy , one recovers the factor-50 discrepancy quoted above . )
the arguments above indicate a worrisome discrepancy between the observed jerking motion and the mechanics of folding presumed to account for this motion . the discrepancy is serious ; it survives even when we use conservative bounds in our estimates . in this section
we survey possible ways to account for the discrepancy .
first we review possible flaws in our description of the kinetic energy and the folding forces .
then we consider other forces that might account for the jerking motions .
our estimate of the kinetic energy was a simplified one , but it gives a proper lower bound for a given displacement @xmath4 and time @xmath5 .
we will proceed by reinforcing several of the assumptions .
firstly , we assumed a constant pressure , though the actual unbalanced pressure increases with time .
however other choices would have led to a higher kinetic energy .
we consider the effect of replacing our constant - pressure estimate by allowing the pressure to increase with time . in order to achieve the required @xmath95 in the given time @xmath5 with a time - increasing pressure
, we will necessarily reduce @xmath96 at early times and increase it at late times
. however , any shift of @xmath96 from earlier to later times has the effect of increasing the kinetic energy . to see this
, we consider a small decrease of @xmath96 at time @xmath97 over a brief interval @xmath98 . to maintain a fixed total displacement @xmath4
, we make an equal addition to @xmath96 at a later time @xmath99 .
any shift of @xmath96 from earlier to later times can be accomplished by repeating this process .
this perturbation of @xmath96 creates a corresponding perturbation of the fluid velocity at the final time @xmath5 : we denote it by @xmath100 .
we may then express the final kinetic energy @xmath101 in terms of this @xmath102 and the initial profile @xmath103 using the integral of eq .
[ solution_real ] .
@xmath104 ^ 2 \nonumber\\ & & = e_{0k}(t)/w + l \rho \int dz~ v_0(z , t)~ \delta v(z , t ) ~~+ { \cal o } ( \delta v^2 ) \label{kineticdeltav}\end{aligned}\ ] ] the second perturbing term is necessarily positive provided the ( positive ) @xmath105 profile is monotonic . to see this we express @xmath102 in terms of the green function @xmath106 of eq .
( [ g ] ) : @xmath107.\ ] ] we may express any monotonic @xmath108 as a sum of positive step functions extending from 0 to some @xmath109 . for a given step function , the contribution to @xmath86 is given by @xmath110 .
thus it suffices to show that this integral is positive . in terms of the @xmath106 functions ,
this means @xmath111 is an increasing function of @xmath112 .
this may be verified explicitly using eq .
( [ g ] ) .
thus a small shift in the pressure profile from a constant one to an increasing one with the same @xmath4 and @xmath5 only increases the energy @xmath86 .
if further small shifts are added , the same reasoning implies that the @xmath86 again increases , provided the starting @xmath105 remains monotonic in @xmath46 .
we conclude that the constant - pressure @xmath86 of eq .
( [ max_kinetic ] ) under - estimates @xmath86 .
a second assumption that affects our estimate of the kinetic energy is the no - slip boundary condition between the sliding monolayer and the fluid beneath . having a slip would cause less drag of the fluid and thus less kinetic energy .
but in our case of hydrophilic heads facing the water , there is no justification for a significant slip . a further effect that can potentially reduce the kinetic energy is the possibility that the motion is restricted in area , so that less fluid is entrained .
as noted above , this restriction can occur if the surface layer is compressible .
the expected compressibility is such that there could be a noticeable effect on the scale of the entire @xmath1 cm sample . to account for this
, we assumed that the motion is restricted to a range of only 5 cm . on this scale we argued that any departure from rigid sliding of the sheet would be negligible .
one possibility we did not consider is that the modulus of the sheet , @xmath8 , is smaller than observed in experiments ( that dealt with sheets without any folds ) .
the existence of folds would make the material weaker , softer .
if that is indeed the case , the part of the sheet that moves rigidly due to jerking could be smaller .
however , in order to resolve the discrepancy , the rigid moving part must be a few millimeters at most , which is not consistent with the statistics of observed jerks @xcite . another potential way to resolve
this discrepancy is that we have _ under_estimated the folding pressure @xmath24 .
as was noted in section [ sec_dynamics ] , the creation of a fold reduces the pressure .
since folding must occur when the pressure exceeds @xmath113 no greater excess pressure can be sustained .
this is true even if additional folds are present .
could it be that @xmath114 mn / m ? this would imply that we have greatly underestimated the bending rigidity @xmath36 .
the local structure of the monolayer might imply a bending rigidity greater than is usually observed due to texture in the sheet @xcite , but how much larger ? equating the folding and kinetic energies requires @xmath34 of at least @xmath115 i.e. @xmath116 j a suspiciously larger value .
we do not see how to resolve the kinetic discrepancy discussed here without departing qualitatively from the wrinkle - to - fold model of ref .
@xcite . to resolve the discrepancy , either the kinetic energy must be smaller than our conservative bounds or
the driving force must be stronger than our bounds .
the former possibility seems unlikely .
our estimates for the kinetic energy are based on direct observations of the motion , together with simple and unquestioned hydrodynamics . however , our account of the driving force depends explicitly on the mechanical properties of a folding monolayer leading to the pressure @xmath34 that we estimated .
perhaps this mechanical picture is wrong .
what other forces are strong enough to trigger the observed motion ?
one possible force is the force of adhesion between two folds that touch .
these forces are comparable to the surface tension of the fluid and are thus many times larger in magnitude than the pressure @xmath34 .
if adhesion forces are responsible , then the observed jerks must take place only after the folds have touched .
this leaves unexplained how the folds came to touch and what causes the jerking to stop .
a second way to have a larger force is to abandon our picture of a simple molecular monolayer .
the observed jerks occur when many folds are already present .
the out - of - plane structure from these prior folds could well impart great rigidity to the surface layer and increase its buckling pressure by a large factor .
if such structures were important , it would qualitatively alter our picture of how the jerks occur .
there would no longer be a clear connection to the simple wrinkle - to - fold model that gave a plausible account of the jerk displacements .
the force responsible for jerks in the monolayers of ref .
@xcite may well be important in a broader context .
the wrinkle - fold transition has been implicated in a broader class of nanoscale systems : nanoparticle trilayers @xcite and single - component lipid monolayers @xcite .
the various energies in these systems are different , and the motion is also somewhat different from that of ref .
nevertheless , the discrepancy shown above may also apply to these other systems .
it is also possible that the force responsible for jerks affects structure as well as dynamics .
that is , the magnitude of a fold may be governed by other characteristic lengths than the wrinkle wavelength @xmath0 .
still , the incipient instability might be due to the wrinkle - fold mechanism while the rapid subsequent motion is controlled by adhesion or some other force .
this work highlights the distinctive dynamics of thin solid sheets on fluids in general .
equation ( [ power32 ] ) applies whenever such a sheet is accelerated by an unbalanced force , whether folding occurs or not .
the equations imply a distinctive form of acceleration and a distinctive partitioning of the supplied work into kinetic energy and dissipation .
the wrinkle - to - fold mechanism @xcite for thin film buckling has allowed a new avenue for understanding buckling phenomena in a range of nanoscale systems .
its successes in explaining structural aspects of folding have led us to apply it to the well - studied dynamics of monolayer jerks . for these dynamic phenomena
, our study indicates that other forces are at play .
supporting this conclusion are experiments on lipid - coated micro - bubbles , where it is clear that gravity , an important ingredient of the folding mechanism , plays no role @xcite .
these other forces must be much stronger than those previously considered .
the necessity of such forces underscores the remarkable nature of monolayer jerks .
it also raises the importance of understanding these jerks .
we are grateful to stuart rice , benny davidovitch , luka pocivavsek , and philippe guyot - sionnest for useful discussions .
this work was supported by the us
israel binational science foundation ( grant no .
2006076 ) , and the university of chicago mrsec program of the nsf under award number dmr 0820054 .
g. kretzschmar , j. li , r. miller , h. motschmann , and h. mhwald , colloid surf .
a * 114 * , 277 ( 1996 ) .
j. krgel , g. kretzschmar , j. b. lee , g. loglio , r. miller , and h. mhwald , thin solid films * 284285 * , 361 ( 1996 ) . j. b. lee , g. kretzschmar , r. miller , and h. mhwald , colloid surf . a * 149 * , 491 ( 1999 ) .
this @xmath117 factor is not to be confused with the energy difference mentioned in the abstract .
in particular changing our choice of bending stiffness would not resolve the energy discrepancy noted in the abstract .
it would worsen this discrepancy .
h. diamant , and t. a. witten , phys .
lett . * 107 * , 164302 ( 2011 ) . | we re - examine previous observations of folding kinetics of compressed lipid monolayers in light of the accepted mechanical buckling mechanism recently proposed [ l. pocivavsek _ et al .
_ , _ soft matter _ , 2008 , * 4 * , 2019 ] . using simple models , we set conservative limits on a ) the energy released in the mechanical buckling process and b ) the kinetic energy entailed by the observed folding motion .
these limits imply a kinetic energy at least thirty times greater than the energy supplied by the buckling instability .
we discuss possible extensions of the accepted picture that might resolve this discrepancy . |
linear optics is a useful tool for constructing quantum circuits by utilizing optical interference @xcite .
a number of quantum gates @xcite were demonstrated with linear optics . however , the operation of these gates has been non - deterministic , because deterministic operation requires nonlinear interactions between photons .
the success probability of the linear optical cnot gate is at most 1/4 @xcite , for instance .
therefore , it is quite difficult to implement large - scale quantum computations with linear optics such as shor s factorization algorithm@xcite , because the more quantum gates the quantum algorithm requires , the less the success probability .
it is becoming more important to implement deterministic quantum operations .
we have found that the quantum leader election ( qle ) protocol can be operated deterministically only with linear optics , and have demonstrated nearly deterministic operation in an experiment with two - photon interference .
before considering the qle protocol , let us briefly introduce distributed computing .
a distributed computing system is a set of computers connected via network channels and it offers the fault - tolerance and the scalability of the system .
the system operates autonomously with a distributed algorithm@xcite that effectively coordinates the computers . in a lot of distributed algorithms ,
computers generate a @xmath0 , which is a kind of script that prescribes the operation or gives permission for them to access shared data .
a computer does some operations based on the token and relays it to the next computer .
then , the next computer receives the token and does operations , and so on . after going round the whole network ,
one distributed algorithm is completed .
if the token is lost during the operation , a unique computer must be specified to restore the only one token , otherwise two or more tokens that tangles the operation may be generated .
such a unique computer is called the leader , and the decision of the leader is known as the leader election problem @xcite . since only one leader should be chosen before performing any distributed algorithms , the leader election problem is one of the most fundamental problems in the distributed computing .
many studies on the leader election problem have been done in the long history of computer science .
we can summarize the leader election problem as : leader election problem : : @xcite + _ let n be the number of computers connected to the network .
each computer i @xmath1 has a state @xmath2 and is initially set to identical values @xmath3 .
set the unique computer @xmath4 , to @xmath5 , otherwise , @xmath6
@xmath7 _ in this problem , there are no third persons such as human beings to determine the leader ; therefore , the computers must do the same operations during the protocol .
if the network is non - anonymous , ( @xmath8 for @xmath9 ) , in which all computers have an identifier ( i.e. , mac address ) , this protocol can be solved quite easily by comparing each identifier .
the computer that has the largest value of identifier can be elected to the leader .
however if the network is anonymous ( @xmath10 ) , in which the computers have no identifiers and they are in an identical initial state , the leader election is no longer trivial .
they have to distinguish themselves and finally obtain different identifiers to distinguish one another using the same operation .
in other words , they have to obtain asymmetric states from the symmetric state after identical operations .
it is widely recognized @xcite that no classical algorithm can offer a deterministic solution within a finite time to this problem .
the possible classical algorithms are stochastic @xcite , where each computer automatically generates an identifier at random .
then , some computers generate the same identifiers with a certain probability , and leader election fails .
in contrast , tani @xmath11 proposed a new quantum algorithm @xcite that could solve this problem deterministically within a finite time .
the process for the quantum protocol is as follows : quantum leader election protocol : : + computers with identical quantum states ( @xmath12 ) are connected via quantum and classical communication channels .
a unique leader is elected after all computers repeats the following processes for n-1 times for n parties : + 1 .
entangled state generation , 2 . quantum communication , 3 . local quantum operation , and 4 .
measurement and classical communication .
the quantum state of each computer is identical in this protocol .
after quantum communication , the computers build a quantum correlation by sharing entanglement .
they can finally derive a non - identical measurement result .
we have demonstrated the two - party qle protocol , which makes the implementation simple , because it only requires one execution of processes ( i ) @xmath13 ( iv ) without any ancillary qubits . and
@xmath14,width=321 ] there is a schematic of the two - party qle circuit is shown in fig.[cir ] . since all parties must perform exactly the same operations , alice s setup is line symmetric to bob s .
both alice and bob have two optical ports defined as the @xmath15 and the @xmath14 , respectively , and the four detectors a1 , a2 , b1 , and b2 are named according to the symmetry shown in fig.[cir ] here , the qubit is represented by the photon number , where single photon state @xmath16 represents qubit 1 and vacuum state @xmath17 represents a qubit 0 .
the circuit consists of four parts ( i ) @xmath13 ( iv ) , carrying out the qle process are given as follows : [ [ entangled - state - generation ] ] entangled state generation + + + + + + + + + + + + + + + + + + + + + + + + + + + alice and bob must prepare identical symmetric state @xmath18 , which corresponds to a single photon pair emerged in their guest ports .
note that photons must be indistinguishable with each other , otherwise we would contravene the identity of initial state .
then single photon entangled states are generated at alice and bob by the 50:50 beam splitters ( bs1 ) as shown in fig.[cir ] .
@xmath19 [ [ quantum - communication ] ] quantum communication + + + + + + + + + + + + + + + + + + + + + + these entangled states generated in ( i ) are shared by alice and bob after they exchange their guest ports @xmath20\notag\\ \xrightarrow{\text{\tiny exchanging}}\ket{\psi_{\rm ii}}\equiv & \frac{1}{2}\bigl[\ket{1_h0_g}_\text{a}\ket{1_h0_g}_\text{b}+\ket{1_h1_g}_\text{a}\ket{0_h0_g}_\text{b}\notag\\ & \hspace{1.5pc}+\ket{0_h0_g}_\text{a}\ket{1_h1_g}_\text{b}+\ket{0_h1_g}_\text{a}\ket{0_h1_g}_\text{b}\bigr].\end{aligned}\ ] ] [ [ local - quantum - operation ] ] local quantum operation + + + + + + + + + + + + + + + + + + + + + + + + a @xmath21-shift in the guest ports results in @xmath22 where @xmath23 represents the bell states , which is defined as @xmath24 $ ] and @xmath25 represents the existence of two photons in alice ( bob ) .
[ [ measurement ] ] measurement + + + + + + + + + + + + the @xmath26 in eq . indicates that the state is no longer symmetric between alice and bob .
then , they obtain different measurement results by distinguishing @xmath23 and @xmath27 states from the others by two - photon interference .
we combine the host and the guest ports with a beam splitter ( bs2 ) as shown in fig.[cir ] .
the beam splitters transform the @xmath26state into @xmath28\notag \\ & + \frac{1}{2}\bigl [ \underbrace{\ket{1_\text{a1}0_\text{a2}0_\text{b1}1_\text{b2}}}_{\uparrow \ket{\psi^+}_\text{a}\ket{\psi^-}_\text{b } } + \underbrace{\ket{0_\text{a1}1_\text{a2}1_\text{b1}0_\text{b2}}}_{\uparrow \ket{\psi^-}_\text{a}\ket{\psi^+}_\text{b}}\bigr].\label{eq4}\end{aligned}\ ] ] .expected results of leader election for all expected combinations of four detectors(a1,a2,b1 , and b2 ) :
( i ) either alice or bob received two photons .
( iii ) alice detects photon at detector a1 and bob detects it at detector b2 and vice versa . [ cols="^,^,^,^,^,<,^",options="header " , ] + the number of single counts ( i ) that we observed in the experiment can be written as @xmath29 where @xmath30 indicates the number of generated photon pairs soon after the nonlinear crystal .
single counts consist of both one - photon detection , and two - photon detection and the number of two - photon detections can be written as @xmath31 + then , we can estimate the number of two - photon detections @xmath32 by using the value of @xmath33 measured in a different experiment as @xmath34 equation implies that the actual contribution of two - photon detections in single counts can be 1% .
the corrected coincidence counts are plotted in fig.[baro](a ) .
the theoretically predicted values from table [ table3 ] have been shown in fig.[baro](b ) for comparison .
figure[baro](a ) shows good agreement with ( b ) . bars ( i ) and ( iii ) in fig.[baro ] indicate corrected two - photon detection and @xmath23 detection , respectively . in these cases leader election
is successfully achieved .
bars ( ii ) and ( iv ) , on the other hand , result from imperfect two - photon interference .
however , bars ( ii ) can be regarded as successful events because the detection of alice and bob is different .
bars ( iv ) indicates that the same detector ( a1&b1 , and a2&b2 ) clicked on alice and bob , then leader election failed . hence , the imperfect two - photon interference is responsible for the performance of the experimental qle protocol .
since we focus only on the two photon events , the success probability per trial of qle circuit can be calculated by : @xmath35 where @xmath36 represent the counts of ( i)@xmath13(iv ) . to calculate the success probability from eq .
, we must derive the number of two - photon detections n@xmath37 . since the n@xmath37 counts are obtained from rough estimates of losses as previously
, we can not precisely calculate the success probability .
therefore , we rewrite eq . using the coincidence probabilities between four detectors , derived in appendix [ app1 ] , as @xmath38\label{eq43}\\ & \hspace{3.2 cm } \gamma\equiv r / t \notag\end{aligned}\ ] ] where @xmath39 and @xmath40 correspond to the branching ratio of beam splitters and the visibility of two - photon interference . using eq .
, we can evaluate performance of qle circuit with the imperfection of two - photon interference , which is characterized by visibility @xmath40 . to estimate the visibility @xmath40 , we observed the two - photon interference fringes shown in fig.[fringe ] .
we recorded two patterns of coincidence counts : ( iii ) a1&b2 , and ( iv ) a2&b2 , which correspond to success and failure , respectively , as a function of the voltage applied to pzt .
both coincidence counts fit the sinusoidal curves with visibility @xmath41 .
hence , the success probability of leader election is calculated as [email protected]% , and the operation is therefore almost deterministic .
let us examine how advantageous the experimental results were over the classical limit . to do this
, we should not consider only success probability but also communication complexity .
communication complexity is characterized by the costs of communications between computers to complete algorithms .
it is commonly recognized in computer science that performance of distributed algorithms can be evaluated with the communication complexity . as shown in appendix [ app2 ] , the optimal classical algorithm involves repeating and comparing the one - bit random number generated by both alice and bob until they generate a different number .
the expected minimum costs of communication @xmath43 , to complete leader election is calculated as 4 . the quantum leader election , on the other hand , is completed by operating the quantum circuits shown in fig .
as noted in sec.ii , alice and bob can finish leader election without the classical communication .
the total communication costs between alice and bob is 2-qubit quantum communication to exchange their guest ports .
it is clear that ideal quantum leader election is advantageous over classical algorithms because leader election is completed deterministically only with 2-qubit quantum communication .
however , the two photon interference is not perfect in the experiment .
that causes a failure of qle with a finite probability ( [email protected]% per one trial ) . in the circuit depicted by fig2 ,
the detection pattern for alice and bob become the same ( ( iv ) a1&b1 or a2&b2 ) with certain probability . then
, if alice receives a single photon at `` a1 '' , bob might receive one at `` b1 '' .
she can not determine at which detector bob has actually received a single photon . since the leader must be elected uniquely , only the successful detection event ( iii ) must be _ post - selected_. therefore , alice and bob have to inform one another of their detection events using 2-bit classical communication to verify their detection events .
the expected communication costs can be calculated along the flow chart given in fig.[flow ] : alice and bob check the photon number they received .
if they received two photons with a probability of @xmath44 , then the leader election is completed only with 2-qubit quantum communication .
otherwise , they need to verify the results with 2-bit classical communication .
if their detection event is ( iii ) with a probability of @xmath45 , leader election is completed with a total of 4-bit communication , otherwise it fails with a probability of @xmath46 .
the probability of failure until the @xmath47 - 1 th trial is written by @xmath48 and the amount of communication to complete leader election at the @xmath47-th trial before and after the verification process are written as @xmath49 and @xmath50 .
then the expected communication costs in qle protocol @xmath51 is calculated as : @xmath52p^{(2)}_{suc}(1-p_{suc})^{k-1}\notag\\ & \hspace{1pc}+\sum_k^\infty\bigl[4(k-1)+4\bigr]p^{(1)}_{suc}(1-p_{suc})^{k-1}\label{eqq1}\end{aligned}\ ] ] we calculated eq . by substituting eq . and and the @xmath51 is ( in bits and qubits ) of quantum leader election as a function of the interference visibility @xmath40 of alice and bob s photons .
, width=302 ] plotted in fig.[complot ] as a function of two - photon interference fringe visibility @xmath40 .
note that without two - photon interference ( @xmath40=0 ) the @xmath51 is equivalent to the classical limit .
however , as shown in fig.[complot ] even if two - photon interference is perfect ( @xmath40=1 , @xmath53=1 ) @xmath51 does not become that of the ideal case , but approaches to three .
this extra one - bit is still leveraged for the verification of the detection events , because alice and bob never know that the operation is in the ideal regime .
we obtain a visibility @xmath40 of 0.845@xmath54% , the expected total communication costs can be calculated as [email protected] which overcomes the classical limit . finally , we touch upon the extension of our implementation for more than three party case . in the case of multi parties ,
the qle protocol requires cnot operation , which is no longer deterministic with linear optics .
moreover all the parties must prepare a cat state and exchange them .
the probability to generate such multi photon states with spdc would be much lower and the optical circuit would be much more complicated than two party case . for the above reasons , it is difficult to implement multi party qle protocol with liner optics .
we demonstrated the experimental 2-party quantum leader election . unlike conventional probabilistic quantum circuits based on linear optics ,
our leader election circuit can operate deterministically only with linear optics .
we derived the success probability @xmath55 and expected total communication costs to complete leader election @xmath51 in terms of two - photon interference visibility in order to evaluate our experimental result .
finally , we have obtained nearly deterministic results for a success probability of [email protected]% and an expected total communication costs of [email protected] , both of which improve upon the classical limit .
we would like to thank dr .
k. matsumoto and dr .
m. hayashi for their helpful discussions , and especially k. yoshino for his kind suggestion and advice on the experiments .
we calculate the coincidence probabilities between four detectors by calculating the overlap integral of field operators .
the field operator just before detectors a1 , a2 , b1 , and b2 can be written as follows @xmath56 where @xmath57 and @xmath58 are the reflectance and the transmittance of four identical beam splitters ( bsa1 , bsb1 , bsa2 , and bsb2 ) , @xmath59 are the time delay of each optical paths , and @xmath60 are the field operators of down - converted photons .
the field operator , @xmath61 , can be written by the fourier transform of the creation operator . @xmath62
the probability of coincidence detection between detector @xmath63 is given by @xmath64 where @xmath65 indicates the wave function of the input state , which is defined @xcite as @xmath66 where @xmath67 is the center frequencies of the pump beam .
the spectral broadening of the pump and the down - converted photons are introduced as @xmath68 and @xmath69 , respectively . by combining eqs.@xmath13 , and after a straightforward algebra , we obtain the probabilities of two - fold coincidence and two - photon detection as follows : @xmath70\label{eqf}\\ p^{(2)}_{b1}=&p^{(2)}_{b2}=\gamma^2t^4\biggl[1+e^{-\frac{\delta\omega^2}{2}\delta
t_b^2}\biggr]\label{eqb}\\ p^{coin}_{a1,a2}=&\gamma(1+\gamma^2)t^4\biggl[1-\tfrac{2\gamma}{1+\gamma^2}e^{-\frac{\delta\omega^2}{2}\delta t_a^2}\biggr]\label{eqa}\\ p^{coin}_{b1,b2}=&\gamma(1+\gamma^2)t^4\biggl[1-\tfrac{2\gamma}{1+\gamma^2}e^{-\frac{\delta\omega^2}{2}\delta t_b^2}\biggr]\\ p^{coin}_{a1,b1}=&(1+\gamma^4)t^4\biggl[1+\tfrac{2\gamma^2}{1+\gamma^4 } e^{-\frac{\delta \omega_p^2}{32}(\delta t_a+\delta t_b)^2}\notag\\ & \hspace{2pc}\times\hspace{1pc}e^{-\frac{\delta\omega^2}{8}(\delta t_a-\delta
t_b)^2}\cos\tfrac{\omega_0}{2}(\delta t_a+\delta t_b)\biggr]\end{aligned}\ ] ] @xmath71\\ p^{coin}_{a1,b2}=&(\gamma+\gamma^3)t^4\biggl [ 1-\tfrac{2\gamma^2}{\gamma+\gamma^3}e^{-\frac{\delta \omega_p^2}{32}(\delta t_a+\delta t_b)^2}\notag\\ & \hspace{2pc}\times\hspace{1pc}e^{-\frac{\delta\omega^2}{8}(\delta t_a-\delta t_b)^2}\cos\tfrac{\omega_0}{2}(\delta t_a+\delta t_b)\biggr]\\ p^{coin}_{a2,b1}=&(\gamma+\gamma^3)t^4\biggl [ 1-\tfrac{2\gamma^2}{\gamma+\gamma^3}e^{-\frac{\delta \omega_p^2}{32}(\delta t_a+\delta t_b)^2}\notag\\ & \hspace{2pc}\times\hspace{1pc}e^{-\frac{\delta\omega^2}{8}(\delta t_a-\delta
t_b)^2}\cos\tfrac{\omega_0}{2}(\delta t_a+\delta t_b)\biggr]\label{eql}\end{aligned}\ ] ] to obtain eq .@xmath13 , we introduce @xmath72 , which represent the time difference of two ports on alice and bob , and the branching ratio of beam splitters , respectively . here , we have made the following assumptions for simplicity : ( 1)@xmath73 , i.e. , the time difference is the same for alice and bob , since they are forced to do the same operations during the qle protocol
. we also assumed ( 2)@xmath74 .
i.e. , the spectral broadening of the photon pair is half that of pump beam ( @xmath75 ) .
further , we assumed ( 3 ) @xmath76=0 , i.e. , the relative phase between the @xmath15 and the @xmath14 in alice and bob is set to 0 .
therefore , the coincidence probabilities can be approximated as @xmath77,\label{eqap1}\\ p_{a1,a2}=p_{b1,b2}=&\gamma\frac{1+\gamma^2}{\left(1+\gamma\right)^4}\bigl[1-\tfrac{2\gamma}{1+\gamma^2}\nu\bigr],\\ p_{a1,b1}=&\bigl[1-\tfrac{2\gamma^2}{1+\gamma^4}\nu\bigr],\\ p_{a2,b2}=&\frac{2\gamma^2}{\left(1+\gamma\right)^4}\bigl[1-\nu\bigr],\\ p_{a1,b2}=p_{a2,b1}=&\gamma\frac{1+\gamma^2}{\left(1+\gamma\right)^4}\bigl[1+\tfrac{2\gamma^2}{\gamma+\gamma^3}\nu\bigr]\label{eqap2}\end{aligned}\ ] ] where @xmath78 implies the temporal overlapping of two photon wave packets on beam splitters .
using@xmath13 we obtain the probability of two - photon detection @xmath44 and single photon detection @xmath45 as @xmath79 hence , the success probability per trial is calculated as @xmath80\ \ \
\biggl(\xrightarrow{\gamma,\nu=1}1\biggr)\label{eqap5}\end{aligned}\ ] ]
how can we obtain optimal classical algorithms with minimum communication complexity ? we obtain the classical limit , @xmath43 , which corresponds to the minimum expected value of the total amount of communication to finish leader election . the classical algorithm for leader election
is to generate and compare the numbers between alice and bob , and then the party who generated the larger number becomes the leader .
the problem is how to generate the numbers and with how many bits to compare them .
let @xmath81 and @xmath82 be the number of bits and the probability to generate `` 1 '' on each bit ( 1-@xmath82 for `` 0 '' ) , respectively ; then , the success probability of leader election per trial @xmath83 is calculated as @xmath84 then , the optimal condition is obtained by differentiating eq .
as : @xmath85 by solving the eq .
, the optimal probability , @xmath82 , is obtained as @xmath86 .
the optimal success probability is calculated as @xmath87 hence , the expected value for the total amount of communication , @xmath88 , after @xmath81-bit comparison is written as : @xmath89 then , the classical limit is calculated as @xmath90 .
99 e. knill , r. laflamme , and g. j. milburn , nature ( london)409 , 46(2001 ) m. koashi , t. yamamoto , and n. imoto , phys .
a 63,030301(r)(2001 ) j .- w .
pan , s. gasparoni , r. ursin , g. weihs and a. zeilinger , nature ( london ) 423 , 417 ( 2003 ) . t. b. pittman , b. c. jacobs , and j. d. franson , phys . rev .
a 64.062311(2001 ) p.w.shor , ieee computer society press p.124(1994 ) d. angluin and a. gardiner , local and global properties in networks of processors , proc .
12th acm symp .
theory of computing , pp.82 - 331 , 1988 n.a.lynch , distributed algorithms , morgan kaufmann publishers , inc .
m. yamashita and t. kameda , ieee _ trans .
parallel distrib .
_ , 7 ( 1 ) : 69 - 89 , 1996 .
a. itai and m. rodeh .
symmetry breaking in distributed networks . _ information and computation _
, 88(1):60 - 67 , 1990 . s. tani , h. kobayashi , and k. matsumoto , exact quantum algorithms for the leader election problem , stacs 2005 , lncs 3404 , pp .
581 - 592,2005 w. p.
grice and i. a. walmsley , phys .
a 56 , 1627(1997 ) j.d .
franson , phys .
lett , 62 , 2205(1989 ) c. k. hong , z. y. ou , and l. mandel , phys .
lett 59,2044(1987 ) r. ghosh , c. k. hong , z. y. ou , and l. mandel , phys .
a 34 , 3962(1986 ) | linear optics is a promising candidate to enable the construction of quantum computers . a number of quantum protocols gates based on linear optics have been demonstrated . however , it is well - known that these gates are non - deterministic and that higher order nonlinearity is necessary for deterministic operations .
we found the quantum leader election protocol ( qle ) can be operated deterministically only with linear optics , and we have demonstrated the nearly deterministic operation which overcomes classical limit . |
the milky way is a typical bright spiral galaxy .
its disk of stars and gas is surrounded by an extended halo of old stars , globular star clusters and a few dark matter dominated old satellite galaxies . for the past 30 years
two competing scenarios for the origin of galaxies and their stellar components have driven much observational and theoretical research .
eggen , lynden - bell and sandage ( 1962 ) proposed a monolithic collapse of the galaxy whilst searle and zinn ( 1978 ) advocated accretion of numerous proto - galactic fragments .
enormous progress has been made in understanding the structure and origin of the milky way , as well as defining a standard cosmological model for structure formation that provides us with a framework within which to understand our origins @xcite .
hierarchical growth of galaxies is a key expectation within a universe whose mass is dominated by a dark and nearly cold particle ( cdm ) , yet evidence for an evolving hierarchy of merging events can be hard to find , since much of this activity took place over 10 billion years ago .
the origin of the luminous galaxy depends on the complex assembly of its @xmath2 dark halo that extends beyond @xmath3 kpc , and on how stars form within the first dark matter structures massive enough to cool gas to high densities @xcite .
the galactic halo contains about 100 old metal poor globular clusters ( i.e. forbes et al . 2000 ) each containing up to @xmath4 stars .
their spatial distribution falls off as @xmath5 at large radii and half the globulars lie within 5 kpc from the centre of the galaxy @xcite .
there is no evidence for dark matter within the globular clusters today @xcite .
the old stellar halo population has a similar spatial distribution and a total luminosity of @xmath6 @xcite .
the stellar populations , ages and metallicities of these components are very similar @xcite . also orbiting the galaxy
are several tiny spheroidal satellite galaxies , each containing an old population of stars , some showing evidence for more recent star - formation indicating that they can hold on to gas for a hubble time @xcite .
half of the dwarf satellites lie within 85 kpc , have luminosities in the range @xmath7 and are surrounded by dark haloes at least 50 - 200 times as massive as their baryonic components @xcite .
cold dark matter models have had a notoriously hard time at reconciling the observed low number of satellites with the predicted steep mass function of dark haloes @xcite .
we wish to explore the hypothesis that cold dark matter dominates structure formation , the haloes of galaxies and clusters are assembled via the hierarchical merging and accretion of smaller progenitors ( e.g. lacey and cole 1993 ) .
this process violently causes structures to come to a new equilibrium by redistributing energy among the collision - less mass components .
early stars formed in these progenitors behave as a collisionless system just like the dark matter particles in their host haloes , and they undergo the same dynamical processes during subsequent mergers and the buildup of larger systems like massive galaxies or clusters . in a recent study ,
diemand et al .
( 2005 ) used cosmological n - body simulations to explore the distribution and kinematics in present - day cdm haloes of dark matter particles that originally belonged to rare peaks in the matter density field .
these properties are particularly relevant for the baryonic tracers of early cdm structures , for example the old stellar halo which may have originated from the early disruption of numerous dwarf proto - galaxies @xcite , the old halo globular clusters and also giant ellipticals @xcite .
since rare , early haloes are strongly biased towards overdense regions ( e.g. sheth and tormen 1999 ) , i.e. towards the centers of larger scale fluctuations that have not collapsed yet , we might expect that the contribution at @xmath8 from the earliest branches of the merger tree is much more centrally concentrated than the overall halo .
indeed , a `` non - linear '' peaks biasing has been discussed by previous authors @xcite .
diemand et al .
( 2005 ) showed that the present - day distribution and kinematics of material depends primarily on the rareness of the peaks of the primordial density fluctuation field that the selected matter originally belonged to , i.e. when selecting rare density peaks above @xmath9 [ where @xmath10 is the linear theory rms density fluctuations smoothed with a top - hat filter of mass @xmath11 at redshift @xmath12 , their properties today depend on @xmath13 and not on the specific values of selection redshift z and minimal mass m. in the following section of this paper we discuss a model for the combined evolution of the dark and old stellar components of the galaxy within the framework of the @xmath14cdm hierarchical model @xcite .
many previous studies have motivated and touched upon aspects of this work but a single formation scenario for the above components has not been explored in detail and compared with data @xcite .
we assume proto - galaxies and globular clusters form within the first rare peaks above a critical mass threshold that can allow gas to cool and form stars in significant numbers ( typically at @xmath15 ) .
we assume that shortly after the formation of these first systems , the universe reionises , perhaps by these first proto - galaxies , suppressing further formation of cosmic structure until later epochs .
we use the n - body simulations to trace the rare peaks to @xmath8 .
most of these proto - galaxies and their globular clusters merge together to create the central galactic region . in section 3
we will compare the spatial distribution and orbital kinematics of these tracer particles with the galactic halo light and old metal poor globular clusters .
we will see that a small number of these first stellar systems survive as dark matter dominated galaxies .
we will compare their properties with the old satellites of the galaxy in section 4 .
= 15.2 cm we propose that ` ordinary ' population ii stars and globular clusters first appeared in significant numbers at redshift @xmath16 , as the gas within protogalactic haloes with virial temperatures above @xmath17k ( corresponding to masses comparable to those of present - day dwarf spheroidals ) cooled rapidly due to atomic processes and fragmented .
it is this ` second generation ' of subgalactic stellar systems , aided perhaps by an earlier generation of metal - free ( population iii ) stars and by their remnant black holes , which generated enough ultraviolet radiation to reheat and reionize most of the hydrogen in the universe by a redshift @xmath18 , thus preventing further accretion of gas into the shallow potential wells that collapsed later .
the impact of a high redshift uv background on structure formation has been invoked by several authors @xcite to explain the flattening of the faint end of the luminosity function and the missing satellites problem within our local group .
here we use high resolution numerical simulations that follow the full non - linear hierarchical growth of galaxy mass haloes to explore the consequences and predictions of this scenario .
dark matter structures will collapse at different times , depending on their mass , but also on the underlying larger scale fluctuations . at any epoch
, the distribution of masses of collapsed haloes is a steep power law towards low masses with @xmath19 . to make quantitative predictions we calculate the non - linear evolution of the matter distribution within a large region of a @xmath14cdm universe .
the entire well resolved region is about 10 comoving megaparsecs across and contains 61 million dark matter particles of mass @xmath20 and force resolution of 0.27 kpc .
this region is embedded within a larger 90 mpc cube that is simulated at lower resolution such that the large scale tidal field is represented .
figure 1 shows the high - redshift and present - day mass distribution of a single galaxy mass halo taken from this large volume .
the rare peaks collapsing at high redshift that have had sufficient time to cool gas and form stars , can be identified , followed and traced to the present day . because small fluctuations are embedded within a globally larger perturbation , the small rarer peaks that collapse first are closer to the centre of the final potential and they preserve their locality in the present day galaxy .
the strong correlation between initial and final position results in a system where the oldest and rarest peaks are spatially more concentrated than less rare peaks .
the present day spatial clustering of the material that was in collapsed structures at a higher redshift only depends on the rarity of these peaks @xcite .
our simulation contains several well resolved galactic mass haloes which we use to trace the evolution of progenitor haloes that collapse at different epochs .
the first metal free population iii stars form within minihaloes already collapsed by @xmath21 , where gas can cool via roto - vibrational levels of h@xmath22 and contract .
their evolution is rapid and local metal enrichment occurs from stellar evolution .
metal - poor population ii stars form in large numbers in haloes above @xmath23^{-3/2}\,m_\odot$ ] ( virial temperature @xmath24k ) , where gas can cool efficiently and fragment via excitation of hydrogen ly@xmath25 . at @xmath26 , these correspond to @xmath27 peaks of the initial gaussian overdensity field : most of this material ends up within the inner few kpc of the galaxy . within the @xmath28mpc turn - around region , a few hundred such protogalaxies are assembling their stellar systems @xcite .
typically 95% of these first structures merge together within a timescale of a few gyrs , creating the inner galactic dark halo and its associated old stellar population . with an efficiency of turning baryons into stars and globular clusters of the order @xmath29
we successfully reproduce the total luminosity of the old halo population and the old dwarf spheroidal satellites .
the fraction of baryons in dark matter haloes above the atomic cooling mass at redshift 12 exceeds @xmath30 .
a normal stellar population with a salpeter - type initial mass function emits about 4,000 hydrogen - ionizing photons per stellar baryon .
a star formation efficiency of 10% therefore implies the emission of @xmath31 a few lyman - continuum photons per baryon in the universe
. this may be enough to photoionize and drive to a higher adiabat vast portions of the intergalactic medium , thereby quenching gas accretion and star formation in nearby low - mass haloes .
the globular clusters that were once within the merging proto - galaxies are so dense that they survive intact and will orbit freely within the galaxy .
the surviving proto - galaxies may be the precursors of the old satellite galaxies , some of which host old globular clusters such as fornax , whose morphology and stellar populations are determined by ongoing gravitational and hydrodynamical interactions with the milky way ( e.g. mayer et al . 2005 ) .
recent papers have attempted to address the origin of the spatial distribution of globular clusters ( e.g. parmentier and grebel 2005 , parmentier et al . 2005 ) .
most compelling for this model and one of the key results in this paper , is that we naturally reproduce the spatial clustering of each of these old components of the galaxy .
the radial distribution of material that formed from @xmath32 peaks at @xmath26 now falls off as @xmath33 within the galactic halo - just as the observed old halo stars and metal poor globular clusters ( cf . figure 2 ) .
cosmological hydrodynamical simulations are also begining to attain the resolution to resolve the formation of the old stellar haloes of galaxies ( abadi et al .
because of the steep fall off with radius , we note that we do not expect to find any isolated globular clusters beyond the virial radius of a galaxy .
these first collapsing structures infall radially along filaments and end up significantly more flattened than the mean mass distribution .
they also have colder velocity distributions and their orbits are isotropic in the inner halo and increasingly radially anisotropic in the outer part .
material from these rare peaks has @xmath34 at our position in the milky way , in remarkable agreement with the recently measured anisotropy and velocity dispersion of halo stars @xcite .
diemand et al .
( 2005 ) show that the radial distribution of rarer peaks is even more highly biased - thus the oldest population iii stars and their remnant black holes are found mainly within the inner kpc of the galaxy , falling off with radius steeper than @xmath35 .
the observational evidence for tidal stellar streams from globular clusters suggests that they are not embedded within extended dark matter structures today @xcite .
this does not preclude the possibility that the globular clusters formed deep within the central region of @xmath36 dark haloes which have since merged together .
( massive substructure within the inner @xmath37 of galactic mass haloes is tidally disrupted i.e. gihgna et al .
this is what we expect within our model which would leave the observed globulars freely orbiting without any trace of the original dark matter component .
however , it is possible that the most distance halo globulars may still reside within their original dark matter halo .
if the globular cluster is located at the center of the cdm cusp , then observations of their stellar kinematics may reveal rising dispersion profiles .
if the globular cluster is orbiting within a cdm mini - halo then we would expect to see symmetric tidal streams orbiting within the potential of the cdm substructure halo rather than being stripped by the galaxy .
the remaining @xmath38% of the proto - galaxies form sufficiently far away from the mayhem that they fall into the assembling galaxy late ( @xmath39 , about one gyr after the formation of the inner galaxy at @xmath40 ) .
this leaves time to enhance their @xmath41 element ratios from type ii supernovae @xcite .
recent studies including chemical modeling of this process support this scenario ( e.g. robertson et al . 2005 , font et al .
2005 ) .
the proto - galaxies highlighted with boxes in figure 1 are those few systems that survive until the present epoch - they all form on the outskirts of the collapsing region , ending up tracing the total mass distribution as is also observed within the milky way s and m31 s satellite systems .
each of our four high resolution galaxies contains about ten of these surviving proto - galaxies which have a radial distribution that is slightly _ shallower _ than that of the total mass distribution but more concentrated than the distribution of all surviving ( or @xmath8 mass selected ) subhalos ( figures [ fig : z0 ] and [ nr ] ) .
this is consistent with the spatial distribution of surviving satellites in the milky way and in other nearby galaxies in the 2df @xcite and deep2 samples @xcite and with galaxy groups like ngc5044 @xcite .
figure 3 shows the distribution of circular velocities of the local group satellites compared with these rare proto - galaxies that survive until the present day .
the local group circular velocity data are the new data from maccio et al .
( 2005b ) where velocity dispersions have been converted to peak circular velocities using the results of kazantzidis et al .
the total number of dark matter substructures is over an order of magnitude larger than the observations .
reionisation and photo - evaporation must play a crucial role in suppressing star formation in less rare peaks , thus keeping most of the low mass haloes that collapse later devoid of baryons .
the surviving population of rare peaks had slightly higher circular velocities just before accretion ( at @xmath42 , dashed line in figure 3 - see kravtsov et al .
2004 ) , tidal stripping inside the galaxy halo then reduced their masses and circular velocities and they match the observations at @xmath8 . dissipation and tidally induced bar formation
could enable satellites to survive even closer to the galactic centre ( maccio et al .
2005a ) .
likewise to the radial distribution , the kinematics of the _ surviving visible _ satellite galaxies resembles closely the one of the dark matter while the same properties for all the surviving subhalos differ ( figures [ vr ] and [ vt ] ) . within the four high resolution cdm galaxy
haloes our 42 satellite galaxies have average tangential and radial velocity dispersions of 0.70@xmath43 and 0.56@xmath44 respectively , i.e. @xmath45 ( the errors are one sigma poission uncertainties )
. these values are consistent with those of the dark matter particles : @xmath46 , @xmath47 and @xmath48 ; the hint of slightly larger dispersions of the satellites are consistent with their somewhat larger radial extent .
in the inner part our model satellite galaxies are hotter than the dark matter background , especially in the tangential component : within 0.3 @xmath49 we find @xmath50 and @xmath51 .
this is consistent with the observed radial velocities of milky way satellites . for the inner satellites also the tangential motions are know ( with large uncertainties however ) ( e.g. mateo 1998 ; wilkinson & evans 1999 ) and just as in our simple model they are larger than the typical tangential velocities of dark matter particles in the inner halo .
the total ( mostly dark ) surviving subhalo population is more extended and hotter than the dark matter while the distribution of orbits ( i.e. @xmath52 ) is similar @xcite .
for the 2237 subhalos within the four galaxy haloes find @xmath53 , @xmath54 and @xmath55 , i.e. there is a similar velocity bias relative to the dark matter in both the radial and tangential components and therefore a similar anisotropy . in the inner halo the differences between dark matter particles and subhaloes are most obvious : within 0.3 @xmath49 we find @xmath56 and @xmath57 .
subhalos tend to avoid the inner halo and those who lie near the center at @xmath8 move faster ( both in the tangential and radial directions ) than the dark matter particles , i.e. these inner subhalos have large orbital energies and spend most of their time further away from the center ( figures [ nr ] , [ vr ] and [ vt ] , see also diemand et al .
2004 ) .
[ haloes ] .present - day properties of the four simulated galaxy haloes .
the columns give halo name , virial mass , virial radius , peak circular velocity , and radius to the peak of the circular velocity curve .
the virial radius is defined to enclose a mean density of 98.6 times the critical density .
the mass resolution is @xmath20 and the force resolution ( spline softening length ) is 0.27 kpc . for comparison with the milk way these halos were rescaled to a peak circular velocity of 195
km / s . in sph simulations of the same halos we found that this rescaling leads to a local rotation speed of 220 km / s after the baryonic contraction @xcite .
the rescaled virial radii and virial masses are given in the last two columns .
[ cols="<,^,^,^,^,^,^ " , ]
we have a implemented a simple prescription for proto - galaxy and globular cluster formation on to a dissipationless cdm n - body simulation .
this allows us to trace the kinematics and spatial distribution of these first stellar systems to the final virialised dark matter halo .
we can reproduce the basic properties of the galactic metal poor globular cluster system , old satellite galaxies and galactic halo light .
the spatial distribution of material within a virialised dark matter structure depends on the rarity of the peak within which the material collapses .
this implies a degeneracy between collapse redshift and peak height .
for example , 3 sigma peaks collapsing at redshift 18 and 10 will have the same final spatial distribution within the galaxy . however
this degeneracy can be broken since the mass and number of peaks are very different at each redshift . in this example at redshift 18 a galaxy mass perturbation has 700 collapsed 3 sigma halos of mass @xmath58 , compared to 8 peaks of mass @xmath59 the best match to the spatial distribution of globular clusters and stars comes from material that formed within peaks above 2.5 @xmath60 .
we can then constrain the minimum mass / redshift pair by requiring to match the observed number of satellite galaxies in the local group ( figure [ mvst ] ) .
if protogalaxies form in early , low mass 2.5 @xmath60 peaks the resulting number of luminous satellites is larger as when they form later in heavier 2.5 @xmath60 peaks .
we find that efficient star formation in halos above about 10@xmath61 up to a redshift @xmath62 matches these constraints .
the scatter in redshift is due to the different best fit redshifts found in our individual galaxy haloes .
after this epoch star formation should be suppressed in small halos otherwise a too large number of satellites and a too massive and too extended spheroid of population ii stars is produced .
the minimum halo mass to form a protogalaxy inferred from these two constraints corresponds to a minimal halo virial temperature of @xmath63 ( figure [ mvst ] ) , i.e. just the temperature needed for efficient atomic cooling .
this model is general for galaxy formation , but individual formation histories may reveal more complexity . soon after reionisation , infalling gas into the already massive galactic mass halo leads to the formation of the disk and the metal enriched population of globular clusters .
the first and second generation of stars forming in proto - clusters of galaxies will have a similar formation path , but occurring on a more rapid timescale .
we find that the mass fraction in peaks of a given @xmath60 is independent of the final halo mass , except that it rapidly goes to zero as the host halos become too small to have sufficiently high @xmath60 progenitors ( see figure [ fielddwarfs ] and table 4 in @xcite ) .
therefore , if reionisation is globally coeval throughout the universe the abundance of globulars normalised to the halo mass will be roughly constant in galaxies , groups and clusters . furthermore , the radial distribution of globular clusters relative to the host halo scale radius will the same ( see diemand et al .
if rarer peaks reionise galaxy clusters earlier @xcite then their final distribution of blue globulars will fall off more steeply ( relative to the scale radius of the host halo ) and they will be less abundant per virial mass @xcite .
observations suggest that the numbers of old globular clusters are correlated with the luminosity of the host galaxy @xcite .
wide field surveys of the spatial distribution of globulars in groups and clusters may reveal the details of how and when reionisation occurred @xcite .
we thank jean brodie , andi burkert , duncan forbes and george lake for useful discussions and andrea maccio for providing the corrected local group data for figure [ fig : massfn ] prior to publication .
all computations were performed on the zbox supercomputer at the university of zrich .
support for this work was provided by nasa grants nag5 - 11513 and nng04gk85 g , by nsf grant ast-0205738 ( p.m. ) , and by the swiss national science foundation . | the milky way contains several distinct old stellar components that provide a fossil record of its formation .
we can understand their spatial distribution and kinematics in a hierarchical formation scenario by associating the proto - galactic fragments envisaged by searle and zinn ( 1978 ) with the rare peaks able to cool gas in the cold dark matter density field collapsing at redshift @xmath0 .
we use hierarchical structure formation simulations to explore the kinematics and spatial distribution of these early star - forming structures in galaxy haloes today .
most of the proto - galaxies rapidly merge , their stellar contents and dark matter becoming smoothly distributed and forming the inner galactic halo .
the metal - poor globular clusters and old halo stars become tracers of this early evolutionary phase , centrally biased and naturally reproducing the observed steep fall off with radius .
the most outlying peaks fall in late and survive to the present day as satellite galaxies . the observed radial velocity dispersion profile and the local radial velocity anisotropy of milky way halo stars
are successfully reproduced in this model .
if this epoch of structure formation coincides with a suppression of further cooling into lower sigma peaks then we can reproduce the rarity , kinematics and spatial distribution of satellite galaxies as suggested by bullock et al .
( 2000 ) .
reionisation at @xmath1 provides a natural solution to the missing satellites problem . measuring the distribution of globular clusters and halo light on scales from galaxies to clusters
could be used to constrain global versus local reionisation models .
if reionisation occurs contemporary , our model predicts a constant frequency of blue globulars relative to the host halo mass , except for dwarf galaxies where the average relative frequencies become smaller .
[ firstpage ] methods : n - body simulations methods : numerical dark matter galaxies : haloes galaxies : clusters : general , globular clusters |
the extragalactic background light ( ebl ) is the surface brightness of the night sky from all extragalactic sources , integrated out to the first epoch of star formation .
the ultimate goal of studying the ebl is to complete the resolution of olbers s paradox by measuring the spectrum of the ebl , identifying the sources that contribute to it , and accounting for their evolution .
constraints on the optical ebl are particularly important because these bandpasses are sensitive to rest - frame uv flux from galaxies at @xmath6 , when we now believe the bulk of star formation in the universe occurred ( madau et al .
1996 ; connolly et al .
thus , measurement of the spectrum of the optical ebl is a critical test for models of the star formation history of the universe ( see , e.g. , whitrow & yallop 1965 ; partridge & peebles 1967a , b ; fall , charlot , & pei 1996 ; vaisanen 1996 ; madau et al .
models for the sources of the ebl must also correctly predict statistical fluctuations in the ebl , which reflect the surface brightness profiles and clustering of these sources . [ for an extensive review on the optical ebl that predates the hdf , see tyson 1995 . ]
the average surface brightness of the sky from detected galaxies sets a firm lower limit on the mean ebl , @xmath7 ( here and throughout this paper , @xmath8 is an average surface brightness ) .
tremendous progress has been made in detecting faint galaxies , through deep surveys such as the hubble deep field ( williams et al .
1996 ) and the near - ir survey by djorgovski et al .
( 1995 ) , which reach as faint as @xmath9mag and @xmath10mag , respectively . however , it is unclear what fraction of the mean optical ebl the detected galaxies comprise , because the uncertainties in the mean level of the optical ebl are larger than the surface brightness contributed by these galaxies .
in addition , the systematic bias of object detection against relatively diffuse , low surface brightness objects leaves open the possibility that we may have missed a substantial fraction of the optical luminosity density in the universe ( disney 1976 ; mcgaugh et al .
1995 ; dalcanton 1997 ) .
figure 1 shows various observational constraints on the mean ebl .
the inset box plots only the optical limits , using linear axes .
solid symbols indicate lower limits on the ebl in several bandpasses from the average sky brightness contributed by detected galaxies , @xmath11 . in optical bandpasses ,
these lower limits include galaxy counts from the hdf and brighter counts from previous surveys ( compilation by pozzetti et al .
uv galaxy counts are from milliard et al .
open circles with @xmath12 uncertainties show a recent detection of the mean optical ebl ( bernstein 1997 ; bernstein , freedman , & madore 1998 ) .
arrows are upper limits from various experiments ( paresce 1990 ; toller 1983 ; dube , wickes , & wilkinson 1977 , 1979 ; hauser 1996 ) , where the upper bound in each case is set primarily by uncertainties in foreground subtraction
. other bounds on the optical ebl ( not plotted here ) include experiments by roach & smith ( 1968 ) , lillie ( 1968 ) , spinrad & stone ( 1978 ) , boughn & kuhn ( 1986 ) , and mattila ( 1990 ) .
precise measurement of the mean level of the ebl is extremely difficult because several foreground sources , particularly zodiacal light , are at least an order of magnitude brighter than the mean optical ebl . at optical wavelengths , uncertainties in the mean ebl leave room for considerable surface brightness above the lower limits set by galaxy detections .
the @xmath13 uncertainties of the optical ebl detection by bernstein et al .
span the range from @xmath14 to @xmath15 .
the gap between previous upper limits on the ebl and the galaxy surface brightness is of order @xmath16 .
these uncertainties prompt us to examine the plausibility of an additional contribution to the ebl that is comparable to the surface brightness from detected galaxies with magnitude @xmath17mag .
could this flux arise from fainter galaxies ?
the surface brightness per magnitude depends on the number counts as @xmath18 .
if we extrapolate the galaxy number counts to infinitely faint limits , assuming the logarithmic count slope @xmath19 at the hdf limit , then all galaxies with @xmath20mag would contribute only an additional @xmath21 to the ebl .
if , strangely , the number counts turn up to @xmath22 beyond @xmath9mag , then this contribution rises to @xmath23 .
thus , it seems unlikely that galaxies fainter than the hdf limit contribute significant flux .
however , this does not rule out objects that evade direct detection , e.g. , galaxies with total magnitude @xmath24 and very low central surface brightness , or a diffuse intergalactic component . if a substantial fraction of the ebl resides in sources that now lie below the detection limits of deep optical imaging , this population might be inferred from its statistical signature in the object - masked `` sky . ''
the internal profiles of sources and clustering among them will cause correlations in the sky brightness .
differences between the spectra of these sources and the average sky brightness will yield corresponding fluctuations in color ( for hst observations , the `` average sky '' is primarily zodiacal light , for ground - based observations there is also a strong atmospheric component ) . to constrain possible sources of additional ebl ,
we compute the angular autocorrelation function of the unresolved light in the hubble deep field . in other words
, we constrain additional sources of ebl by studying the correlation properties of the flux that can not be assigned with large statistical significance to individual objects .
fluctuation analysis of the sky brightness is complementary to direct measurement of the mean ebl ; although the uncertainties in direct measurement would permit a large ebl component in addition to detected galaxies , the small amplitude of the sky fluctuations rules out many plausible candidates for such a component .
for example , we can statistically test the proposition that previously undetected low surface brightness galaxies comprise most of the ebl ( vaisanen 1996 ) .
this approach is insensitive to a truly uniform optical background , but the most plausible source of the ebl is emission from stars , which are likely to be clustered . at large redshift
, we observe these in the rest - frame uv , where most of the stellar flux is from o and b stars .
regions of active star formation are extremely clumpy , so we expect the flux to be strongly correlated .
the approach of studying the correlation properties of the unresolved optical background was pioneered by shectman ( 1974 ) , who used schmidt plates to detect optical ebl fluctuations on several arcminute scales .
shectman detected an ebl power spectrum that was consistent with clustering of galaxies fainter than @xmath25 , as predicted by gunn ( 1965 ) and shectman ( 1973 ) .
tyson ( 1988 ) found evidence for residual surface brightness fluctuations in deep b band ccd images after cleaning detected @xmath26mag galaxies and smoothing on a scale of @xmath27 .
cole , treyer , & silk ( 1992 ) examined how fluctuation analysis of deep optical images might constrain faint galaxy populations .
martin & bowyer ( 1989 ) detected arcminute - scale fluctuations with a rocket - borne uv detector , which indicated the presence of a uv background from galaxies .
boughn , saulson , & uson ( 1986 ) measured the beam - to - beam variance of the sky flux at @xmath28 m to set upper limits on galaxy counts in the infrared , assuming a model for the galaxy clustering .
jenkins & reid ( 1991 ) also measured sky fluctuations at @xmath28 m to test galaxy evolution models .
recently , kashlinsky et al . (
1997 ) applied a fluctuation analysis to the dirbe sky maps to set constraints on the infrared background .
fluctuations at zero lag have long been used to constrain the number count distribution of radio sources ( scheuer 1957 ; hewish 1961 ; condon 1974 ; see , e.g. , fomalont et al .
1988 for a recent result ) .
hasinger et al .
( 1993 ) computed limits on the x - ray flux distribution in deep rosat images and several analyses have been performed to measure the correlation function of the x - ray background , as observed by heao 1 ( persic et al .
1989 ) and einstein ( barcons & fabian 1989 ) .
in this paper we use the hubble deep field to probe the correlation structure of the night sky after masking all detected objects and thereby establish constraints on possible sources of the optical ebl .
section 2 presents estimates of the auto and cross correlation of the unresolved sky in the hubble deep field .
section 3 discusses possible contributions to these measurement from foregrounds and instrumental systematics . in section 4
we use these results to set stringent limits on the types of additional galaxy populations that might lie hidden below current detection limits and examine if some of the measured correlation signal may be caused by weakly clustered intergalactic light .
section 5 reviews our conclusions .
we quote all hdf magnitudes in the ab system ( oke & gunn 1983 ) , defined as @xmath29 , where @xmath30 is in units of erg / s/@xmath31/hz .
distances are in @xmath32mpc , where the hubble constant is @xmath33 km s@xmath34 mpc@xmath34 .
we analyze the drizzled version 2 images of the hdf observed through the wide field cameras ( wf ) in the f450w , f606w , and f814w filters ( hereafter we refer to these bandpasses as @xmath35 , @xmath1 , and @xmath2 , respectively ) . except for corrections to the flat field calibration ( see below ) , we use the data as reduced and processed by the hdf team ( williams et al .
we analyze only the central one - quarter of the area ( roughly @xmath36 ) of each drizzled image , to minimize the possible impact of residual large - scale gradients , and structure in the dark counts due to fluorescence of the mgf@xmath37 window ( burrows et al .
[ for further details of the hdf , see http://www.stsci.edu/ftp/science/hdf/hdf.html .
] we apply corrections to the version 2 hdf images to remove large - scale gradients that are apparent after object - masking and smoothing of the remaining sky .
comparison of these features with ratio images formed from the v2 flat field calibrations and more recent calibration files ( the latter are the calibration files currently used in the wpfc2 pipeline ) indicate that the gradients are caused by errors in the v2 calibrations . to remove these features we smooth the flat field ratio images ( to remove small - scale noise ) and multiply them into the hdf images .
these gradients are quite small , typically @xmath38 across an entire wf image , and are not expected to affect photometry of individual objects , because algorithms such as focas estimate the sky level near each object .
we remove these gradients because we are interested in detecting r.m.s .
fluctuations in the sky at the level of @xmath39 . in section
we discuss the possible contribution of residual flat field uncertainty to the apparent sky fluctuations . to prepare an object - masked image , we use the hdf version 2 focas catalog to identify sources with total magnitudes brighter than @xmath0 ( williams et al .
estimate the hdf to be 80% complete to @xmath40 ) , which includes virtually every object in the catalog . for comparison
, we also derive catalogs from the hdf using the sextractor package ( bertin & arnouts 1996 ) ; we find the mask structure to be relatively insensitive to the choice of catalog and choose the available v2 focas catalog to allow others to reproduce our results .
the same mask , derived from detections in the summed @xmath41 image , is used for all filters . for each object
we flag the pixels that lie within an ellipse that is twice as large as the area used by focas to compute the `` total magnitude . ''
the focas `` total magnitude area '' is always at least twice the area within the detection isophote , so the masked area is at least four times as large as the isophotal detection region .
smoothing of the masked images on a range of scales does not reveal residual `` doughnuts '' around the masked objects .
this masking procedure removes @xmath42 of the pixels in each wf image .
therefore it does not appear that the sky is confusion - limited to the isophotal detection threshold of @xmath43 mag / arcsec@xmath44 that was used for the focas catalog .
the field of view of the wf cameras is well - matched to probing ebl fluctuations caused by the correlation structure of galaxy profiles .
the wf cameras each have a @xmath45 field of view of and @xmath46 pixel scale . at redshift @xmath47
, @xmath48 corresponds to 8.5kpc ( @xmath49 and @xmath50km / s ) . on these angular scales ,
the clustering amplitude of @xmath51 galaxies is @xmath52 ( colley et al .
1996 ; villumsen et al .
this weak clustering implies that small - scale fluctuations in the ebl will be dominated by the surface brightness profiles of faint objects rather than by the clustering among them . for observed surface brightness @xmath53
, we define the autocorrelation function @xmath54 where @xmath55 .
we often refer to the mean surface brightness @xmath8 as the mean `` sky '' brightness ( after masking detected objects , @xmath8 corresponds to the usual definition of a `` sky level '' ) .
we use cartesian coordinates @xmath56 , with @xmath57 , because the relative angles in the hdf satisfy @xmath58 .
the surface brightness autocorrelation may be written as the product of a dimensionless autocorrelation function and the square of the mean sky brightness @xmath59 to prepare an image for correlation analysis , we compute the mean value of the unmasked pixels and subtract this mean from the image , set pixels within the object ellipses to zero , and taper the edges of the image with a cosine bell function . to estimate the autocorrelation function , we compute the fourier transform of the image , square the moduli of the fourier coefficients ( for cross correlations
, we compute the modulus of the products of the images fourier coefficients ) and correct these amplitudes for the effective area of the image after masking and weighting .
we then inverse fourier transform this two - dimensional power spectrum to produce the two - dimensional autocorrelation function .
averaging over all angles yields the function @xmath60 .
computation of @xmath60 by direct summation of the products of the pixel values yields identical results ( but uses significantly more cpu cycles ) .
analysis of test images with known power spectra confirm that we recover the true signal , after correcting for the integral constraint bias .
the integral constraint bias arises because we estimate the mean surface brightness @xmath8 from the image itself .
the implicit assumption that the ensemble average of @xmath61 is identical to the mean within the image is equivalent to assuming that there are no fluctuations of the mean surface brightness between different images .
this bias causes us to underestimate the autocorrelation by the integral average of the true autocorrelation function over the area @xmath62 of the image , @xmath63 where @xmath64 is the true dimensionless autocorrelation function .
figure 2 shows the dimensionless autocorrelation function of the object - masked hdf `` sky '' in each of the three bandpasses . in each case
, we average the results for the wf2 , 3 , and 4 fields .
representative error bars are attached to the @xmath1 curve only and are estimated from the variation among the three fields and within bins of width @xmath65 . near zero lag , the signal is dominated by photon shot noise over the scale of the @xmath46 wf pixels . on scales from @xmath66 to @xmath67 , these autocorrelation functions are well fitted by a power law @xmath68 with @xmath69 .
table 1 summarizes the power law fits and their uncertainties . at @xmath48 ,
the typical fluctuation is @xmath70 , i.e. , an r.m.s .
correlated fluctuation of @xmath71 of the mean sky brightness , or @xmath72 mag / arcsec@xmath44 in @xmath1 .
the turnover of @xmath73 on scales @xmath74 is consistent with the integral constraint bias that we expect for a power law correlation function over a field of this size . using the power law fits in table 1
, we integrate equation ( 3 ) over a @xmath36 field and find that the correlation functions in figure 2 are underestimated by approximately @xmath75 ( note heavy solid bar in figure 2 ) . what could account for the power law shape of these correlations ? in section 3
we discuss several instrumental systematics and foregrounds that might contribute to the measured correlations .
we note that the unmasked wings of the point spread function might cause correlations with similar shape to those plotted in figure 2 .
to explain the shape of the correlation function with extragalactic sources , one can imagine an unclustered population of sources of different scale size such that the sum of their autocorrelations yields a power law .
alternatively , the similarity of the power law index , @xmath69 , to that observed for galaxy clustering , @xmath76 , suggests that the clustering of sources causes the power law shape .
however , because we have reason to believe that systematics affect this measurement , it is premature to fit such models .
if we only mask galaxies brighter than @xmath77mag , rather than @xmath78mag , we find autocorrelations at @xmath79 that are an order of magnitude above the curves in figure 2 .
this test indicates that fluctuation analysis of shallower imaging data would be dominated by the profiles of galaxies detected by the hdf . in ground - based imaging ,
this small - scale correlation structure would be smeared by seeing and the correlation signal from the profiles of galaxies in the magnitude range @xmath80 would be overwhelmingly larger than the signal measured in the hdf .
in other words , the hdf is the first data set in which we could have detected a signal as weak as that shown in figure 2 .
one might worry that , just as correlations in shallower imaging would be dominated by galaxies below the detection threshold , galaxies just fainter than the hdf limit dominate the observed correlations ; in section 4 we show that this is not the case .
cross correlations of images in different filters and comparison of these results with the respective autocorrelations provide a critical test of the nature of the observed fluctuations .
dot - dashed curves in figure 2 show cross correlations of @xmath35 vs. @xmath1 and @xmath1 vs. @xmath2 .
the cross correlations are shown as @xmath81 , where @xmath82 and @xmath83 are the mean sky levels .
the photon count shot noise spike near zero lag is absent in the cross correlations because this source of noise is uncorrelated between the different images .
absence of this spike shows that the power law behavior of the correlations extends down to the smallest scale that we observe .
this smallest scale is @xmath84 rather than the @xmath46 scale of the wf pixels , thanks to the sub - pixel resolution recovered by drizzling ( see fruchter & hook 1997 ) .
the close agreement between auto and cross correlations indicates that similar structure is present in different filters , in exposures that were obtained at different times .
the relative amplitudes are consistent with a common origin for most of the fluctuations in all three filters , with some extra signal in the @xmath35 and @xmath2 images that is not present in @xmath1 ( @xmath1 also has the highest signal - to - noise ratio ) .
the dimensionless correlations in figure 2 represent fluctuations relative to the mean sky level , thus the good match between correlations in different filters implies that the color of the fluctuating component must be close to that of the mean sky .
it is important to remember that , because the mean sky is much brighter than any possible ebl component , an instrumental systematic that varies slowly with wavelength could cause correlated structure with the same color as the mean sky .
we have also computed the autocorrelation function of the color of the unresolved sky , i.e. , the ratio of surface brightness in different bandpasses ( note that this is the ratio of measured surface brightness _ before _ subtracting the mean sky ) .
color correlations of the @xmath85 and @xmath86 ratio images , expressed as @xmath87 , have slightly smaller amplitude and steeper slope than the auto and cross correlations .
color correlations should be relatively less affected by flatfield uncertainties because any wavelength-_independent _ flatfield structure is divided out in forming the ratio image .
we will report in detail on these results in paper ii ( vogeley 1998 ) .
an important point is that we expect the amplitude of color correlations to be smaller than the individual filter autocorrelations by a factor that strongly depends on the relative color of the sources of the fluctuations and the mean sky .
here we note only that the detection of color correlation of the same order as the auto and cross correlations argues against a dominant systematic contribution from flatfield errors .
several instrumental systematics and foreground sources may contribute to the clustering signal plotted in figure 2 .
these effects include errors in calibration of the instrument sensitivity , smearing by the point spread function ( psf ) of the telescope , wide - angle scattered light , and fluctuations in the zodiacal and galactic foregrounds .
because these systematics and foregrounds only add to the measured clustering signal , we may use the observed correlations as a firm upper limit on the true ebl fluctuations ; this is the approach that we follow in section 4 below .
if we identify an instrumental effect that contributes to the fluctuations and subtract this contribution , then we obtain an even stronger constraint on additional sources of the ebl . for example , we have already been able to remove some signal by correction for obvious flatfield errors , and preliminary analysis clearly indicates a contribution from scattering by the telescope psf . in paper ii we provide a detailed analysis of these possible contributions , with the goal of improving these upper limits .
errors in calibration of spatial variations in the instrument sensitivity will cause apparent fluctuations in the sky . as noted above , smoothing of the object - masked sky revealed large - scale features in the v2 hdf images , which we removed using better flat field calibrations .
color correlation analysis removes signal caused by wavelength - independent flatfield errors , but this does not rule out a wavelength - dependent component .
another test is to compare the two - dimensional autocorrelation structure of the flatfields themselves with the detected signal .
if the uncertainties in the flatfields are proportional to the flatfields themselves , then features in their two - dimensional autocorrelations will show up in the sky correlations .
in addition to cross - correlating the flat field images with the data directly , we can compare the multipole moments of their two - dimensional autocorrelation functions .
preliminary analyses do not reveal cross correlations or anisotropies that are characteristic of flat field errors .
measurement of the multipole moments can also reveal problems caused by charge transfer efficiency or other defects aligned with the ccd rows or columns , which cause a quadrapole signal .
scattered light from the hst secondary mirror support spider creates an `` x '' pattern that would yield a large hexadecapole moment .
the point spread function of the wfpc2 has extended wings , which probably arise from scattering inside the cameras ( krist & burrows 1994 ; krist 1995 ; burrows et al . 1995 ) .
this psf is anisotropic and varies with position within each camera . for our purpose ,
the most important effect of the psf wings is to scatter some flux well beyond the mask region of an object .
a crude test for this effect is to make an image in which we place a model of the psf at the position of each detected object , with amplitude proportional to the measured magnitude , then mask this image as we do for the data and cross - correlate with the masked hdf image .
preliminary results from this and similar tests ( such as varying the mask size ) indicate that this effect produces correlations with shape that is similar to the hdf correlations and could contribute as much as @xmath88 of the observed correlation amplitude
. this may be the dominant systematic effect on our measurement .
however , it is difficult to cleanly separate this psf effect .
similar effects would arise if the profiles of detected galaxies extend far beyond the masked regions ( i.e. , if the sky is confusion - limited in these galaxies at some very faint isophote ) , or if lower surface brightness objects are clustered with the detected galaxies .
the dark count rate in the wfpc2 varies both spatially and temporally ( burrows et al .
these variations are probably caused by fluorescence of the mgf@xmath37 window on the camera . at fixed epoch ,
the dark count rate is constant within the central region of each field , but declines near the edges of the ccd .
the amplitude of the dark count rate , and therefore the steepness of the roll - off , varies in time with the cosmic ray flux . to minimize sensitivity to this effect
, we restrict our analysis to the central one - quarter of the area of each ccd .
the drizzling procedure that was used to combine the hdf exposures includes corrections for the geometric field distortion in each camera .
restricting our analysis to the central area of each field also minimizes sensitivity to residuals from this distortion .
the motion of the earth relative to the source of the zodiacal light causes fluctuations in this foreground to be decorrelated between wfpc2 exposures .
the zodiacal light is solar flux that is backscattered from a layer of dust that extends to roughly 3 a.u .
( dermott et al . 1996 ; reach et al .
1996 ) . in the time it takes for hst to orbit the earth ( 96 minutes ) , the earth s motion around the sun changes the hst s line of sight through this dust layer . during the month of december (
when the hdf was observed ) , the line of sight through a screen at a distance of 3 a.u . towards a fixed extrasolar target at @xmath89 @xmath90 ( the position of the hdf ) changes by @xmath91 .
the field of view of a wf camera is @xmath92 , therefore sky brightness fluctuations due to the zodiacal foreground become decorrelated if we examine fluctuations by cross correlating wfpc2 exposures that are separated in time by at least the duration of one orbit .
exposures in different filters were separated by many orbits , thus cross correlations of images in different filters should be unaffected by zl fluctuations .
the good match of auto and cross correlations indicates that zl fluctuations do not strongly affect any of our measurements .
the next brightest foreground is galactic cirrus , which is presumably the reflection of starlight from high - latitude dust .
the hdf field was chosen to lie at a minimum in the iras @xmath93 m maps , thus we expect this galactic foreground to be much smaller than average .
guhatakurta & tyson ( 1989 ) measured the color of the galactic cirrus and found that it is @xmath94 redder in @xmath95 than either the faint blue galaxies or the structure with surface brightness fainter than @xmath96 mag / arcsec@xmath44 seen on scales @xmath97 by tyson ( 1988 ) .
this color is similarly too red to match the color implied by the auto and cross correlations of the hdf sky ( within a few tenths of a magnitude of the zodiacal light color ) .
however , because little is know about correlations of this foreground on scales of a few arcseconds , we can not exclude this possibility .
is there a population of sources that ( 1 ) would not have been directly detected in the hdf , ( 2 ) would contribute significant surface brightness to the ebl , and ( 3 ) would not overproduce fluctuations in the ebl ? if we specify the distribution of fluxes , surface brightness profiles , and angular clustering of a proposed undetected population , then we can compute the effect of these sources on fluctuations in the ebl . using the measured correlations of the object - masked sky in the hdf as an upper limit on the true ebl fluctuations
, we then test whether the predicted fluctuations are allowed by the measured sky correlations .
the autocorrelation of the surface brightness from a population of sources is the sum of contributions from the correlation structure within the profiles and from clustering among the sources . for a population of sources with identical apparent surface brightness profiles @xmath98 , the autocorrelation function is @xmath99 the first term is from the correlations of flux within the object profiles ; @xmath100 is the convolution of a galaxy profile with itself , @xmath101 the second term is from clustering among the galaxies , with angular two - point correlation @xmath102 . in this subsection
we assume that @xmath103 , so that we can constrain possible ebl sources on the basis of their profiles alone .
the addition of source clustering would increase the predicted ebl fluctuations and strengthen the constraints on these models . in the next subsection
we examine the other extreme , a clustered diffuse ebl source . because it does not account for absorption of flux from one galaxy by dust in another , equation ( 5 )
is exact only in the limit where the sources are transparent or sparsely distributed .
if galaxies have finite optical depth then we require more sources to produce fixed optical ebl .
another effect of absorption is that clustering of sources causes a smaller increase in the ebl fluctuations because nearby galaxies tend to shield one another .
a complete accounting for these effects requires specification of the redshift distribution , spectral energy distribution , and extinction law for the sources
. we will examine such detailed models in paper iii .
here we consider only simple phenomenological source models , in which unclustered sources have exponential surface brightness profiles with identical apparent central surface brightness @xmath104 and apparent angular scale length @xmath105 .
such a population contributes total surface brightness @xmath106 where @xmath107 is the projected number density , @xmath108 is the total flux of each source , and @xmath109 is the ratio of this extra surface brightness to the surface brightness in detected galaxies to the hdf limit , @xmath110 ( here we define @xmath110 , which is the same as @xmath11 defined in section 1 , to avoid confusion with the additional source population ) . to compute the ebl fluctuations from this source population ,
we first compute the convolution of an exponential disk with itself , @xmath111 which defines a dimensionless profile function @xmath112 with @xmath113 .
the autocorrelation of the population is @xmath114 the power law fits to the hdf correlations ( table 1 ) yield a constraint on the ebl fluctuations caused by this population , @xmath115 for all @xmath116 . combining the equations above ,
we derive a constraint on the the @xmath1 band central surface brightness @xmath104 and angular scale length @xmath105 of these sources , @xmath117 if we define an effective area @xmath118 for each source , then the covering factor ( the average number of objects along a random line of sight ) is of order @xmath119 figure 3 shows the resulting constraints on the @xmath1 band central surface brightness @xmath104 and scale size @xmath105 of a population of identical unclustered exponential disks that contributes mean surface brightness equal to that of detected galaxies ( @xmath120 ) .
the region below the solid line satisfies the constraint of equation ( 11 ) . above this line
the sources autocorrelation function would exceed this upper limit for some range of @xmath116 .
the right - hand axis indicates the covering factor @xmath121 for different @xmath104 . above the dashed line at the top of this figure lie objects directly detected in the hdf
using focas with an isophotal threshold of @xmath43 mag / arcsec@xmath44 . for a smaller contribution to the ebl ,
i.e. , for @xmath122 , we relax the @xmath104 limit by @xmath123 and change the covering factor by a factor @xmath124 . in the limit @xmath125 , the allowed region abuts the detected region . two examples of populations that contribute @xmath126 and marginally satisfy the correlation constraint illustrate how very different such a population would be from the hdf detections . for @xmath127 a population of extremely faint , total magnitude @xmath128 , sources with @xmath129
mag / arcsec@xmath44 is marginally allowed .
the number of such objects , @xmath130deg@xmath131 , is @xmath132 times the total number per deg@xmath44 of detected galaxies . for @xmath133 ,
the correlations allow a population of @xmath134mag sources , with @xmath135 mag / arcsec@xmath44 . for comparison ,
the typical scale size of @xmath136mag detected galaxies is @xmath137 .
the projected number density of the additional @xmath134mag population would be 20 times the total number of detected objects .
on the righthand axis of this plot we note the covering factor that corresponds to each choice of @xmath104 and @xmath105 . for the former example ,
this covering factor is @xmath138 , for the latter @xmath139 .
such large covering factors imply that these objects might be detected as qso absorption line systems and/or from reddening of background objects . for comparison to these models , note that the most extended low surface brightness galaxy seen to date , malin 1 , has an extrapolated disk central surface brightness of @xmath140 mag / arcsec@xmath44 in v and scale size @xmath141kpc ( bothun et al .
1987 ) . at @xmath142
this would be dimmed to roughly @xmath143 mag / arcsec@xmath44 in i and would have an apparent scale size @xmath144 .
we obtain relatively poor constraints on sources with @xmath145 because the integral constraint bias is comparable to the correlation function on these scales ( see table 1 ) ; the @xmath36 field that we analyze is too small to accurately measure larger scale fluctuations .
however , for @xmath120 , objects like malin 1 clearly would be ruled out in fig . 3 for a reasonable extrapolation of the measured correlation function .
thus , to make a large contribution to the ebl , the correlation constraint requires that this additional population and the detected galaxies form an extremely bimodal distribution in surface brightness .
the undetected sources must have very low central surface brightness and be confusion - limited on the sky .
some galaxies certainly lie within the region between the focas detections and the `` allowed '' region for extra sources of the ebl .
as shown in section 1 , extrapolation of the number counts would place some objects here but they would not contribute very much surface brightness to the ebl .
extrapolation of the detected galaxy counts adds little to the ebl , but would these fainter galaxies cause detectable correlations ? no .
following equation ( 9 ) , a population of galaxies fainter than @xmath146 with number per magnitude distribution @xmath147 and identical angular scale size @xmath105 has zero - lag autocorrelation signal @xmath148 { \mathrm{(erg / s / cm^2/sr / hz)^2}}.\ ] ] if we extrapolate the hdf v - band counts from @xmath9mag to zero flux with logarithmic count slope @xmath19 and scale size @xmath149 , then @xmath150(erg / s/@xmath31/sr / hz)@xmath44 , which translates to a dimensionless correlation @xmath151 . comparison to figure 2 shows that this signal is two orders of magnitude below the measured correlations .
thus , although galaxies brighter than the hdf detection limit would dominate the clustering signal ( see section 2 ) , galaxies just fainter than this limit have little effect on the observed correlations .
so far , we have ignored clustering among the undetected objects .
if the undetected sources are clustered , then the solid line in figure 3 moves down towards even lower allowed central surface brightness . in the limit of point - like objects ,
the autocorrelation function of a population ( eq.[5 ] ) reduces to @xmath152 this must be convolved with the psf to determine the observed autocorrelation .
if we set @xmath153 then we obtain a simple constraint on the @xmath48 angular clustering amplitude of the sources , @xmath154 this constraint applies for all scales if we make the additional assumption that these objects have power law angular correlations , @xmath155 with @xmath69 , similar to that fit to the hdf sky correlations . on scales larger than the psf size
, the inequality in equation ( 15 ) is equivalent to a constraint on a clustered diffuse ebl component with @xmath156 .
figure 4 shows the constraints on combinations of @xmath157 and @xmath158 from the hdf correlations . for @xmath120 , the upper bound on angular clustering of diffuse light is @xmath159
this @xmath48 clustering amplitude is an order of magnitude smaller than the clustering measured for galaxies as faint as @xmath160 in the hdf ( colley et al .
1996 ; villumsen et al .
thus , to contribute as much surface brightness as the detected galaxies , these undetected objects must be not only extremely faint and numerous , but also uniformly distributed to exquisite precision .
apart from _ imagined _ sources of additional ebl , several observations of galaxy clusters detect diffuse light or intergalactic stars that would contribute to the amplitude and clustering of the ebl if similar surface brightness is associated with most galaxies . from r band imaging of abell 2029 , uson , boughn , & kuhn ( 1991 )
infer that 10% of the light in this cluster is in a diffuse component . using wfpc2 imaging of the virgo cluster , ferguson , tanvir , & von hippel ( 1996 )
find a population of intergalactic stars that could contribute total flux equal to 10% of that in galaxies .
theuns & warren ( 1996 ) detect candidate planetary nebulae in fornax and infer an intergalactic stellar population that contributes as much as 40% of the cluster light .
mndez et al . ( 1997 ) also find candidates for intergalactic planetary nebulae in virgo .
tyson , kochanski , & dellantonio ( 1997 ) detect diffuse light in cl0024 + 1654 that comprises @xmath161 of the light in the central 100kpc of this cluster .
if diffuse intergalactic light makes a fractional contribution to the ebl ( relative to @xmath110 from galaxy counts ) that is similar to the intergalactic surface brightness detected in clusters , then it could cause ebl fluctuations close to those measured in the hdf sky .
the constraints on diffuse intergalactic light also follow from equation ( 15 ) if the diffuse light is clustered in similar fashion to faint galaxies , with a power law correlation function @xmath162 .
let us suppose that the clustering amplitude of the diffuse light is comparable to that of faint galaxies .
the small arrow in figure 4 indicates the level of diffuse light that contributes an additional @xmath163 ( i.e. , similar to the @xmath164 contribution of intergalactic stars in virgo ) .
the ebl correlations from this source would match the hdf correlations if this light has the same clustering amplitude as @xmath165mag galaxies , @xmath166 ( couch , jurcevic , & boyle 1993 ) . because @xmath167 of the light from detected galaxies comes from @xmath168 galaxies
, it seems plausible that much of the diffuse light would be associated with galaxies in this same magnitude interval , and be gravitationally clustered with similar amplitude .
note that the contribution of any such a source to the ebl fluctuations would leave even less room in the upper limits on sky correlations for a discrete source population .
a remaining question is whether the hdf is a typical field for the purpose of probing fluctuations in the ebl .
the answer depends on the redshift and angular size of the hypothesized sources of undetected ebl .
the @xmath169 field of view of each wf camera limits the utility of the hdf for studying fluctuations on scales much larger than @xmath4 .
this angular coverage also limits the range of apparent magnitude of the galaxies that the field includes . because this observation was designed to be an unbiased probe of the high - redshift universe , the field was deliberately chosen to avoid bright foreground galaxies ( @xmath170mag ) which would fill much of the wfpc2 field of view .
this avoidance of bright galaxies raises the concern that the hdf is biased against detection of ebl fluctuations from previously undetected sources at redshift @xmath171 that are clustered with bright galaxies .
[ a similar bias arises in the direct ebl detection method of bernstein et al . , because they use wfpc2 imaging to measure the total sky brightness .
] however , it seems extremely contrived to envisage a population of sources that is not detected at @xmath172 through fluctuation analysis , is not directly detected at @xmath173 in deep ground - based imaging , and yet contributes a large fraction of the optical background .
another issue is variation of ebl fluctuations from field to field .
the total imaged area of the three wf cameras is 5.3 arcmin@xmath44 , of which we analyze only the central 1.3 arcmin@xmath44 .
if additional sources of ebl are distributed similarly to the detected galaxies , then expected fluctuations in galaxy counts provides a rough estimate of uncertainty in the mean ebl over this area .
the poisson fluctuation in counts of galaxies with magnitude @xmath174 would be 5% over a 1.3 arcmin@xmath44 area .
for the magnitude range @xmath175 , the expected fluctuation rises to 12% .
galaxy clustering does not significantly increase this lower bound on the field - to - field fluctuations ; although the field of view spans only a few hundred kpc in angle , these magnitude intervals include galaxies over many hundreds of comoving mpc in distance .
we conclude that the field - to - field variation in the amplitude of the hdf correlations is likely to be caused by systematic effects rather than true fluctuations in the ebl .
fluctuations in the object - masked sky in the hdf provide a strong , albeit indirect , constraint on the mean optical ebl .
if the hdf provides a fair probe of the extragalactic sky between detected galaxies , then we conclude that this observation and other deep imaging surveys have already detected the sources of the majority of the optical ebl .
the mean optical ebl is , at most , a few tens of percent above the mean surface brightness from detected galaxies . at @xmath176 , a generous allowance for up to 50%
additional ebl suggests @xmath177erg / s/@xmath31/sr .
this mean level lies at the lower end of the uncertainty range of the measurement by bernstein et al .
( @xmath178erg / s/@xmath31/sr , after including @xmath179mag galaxies ) and well below previous upper limits .
we infer this mean ebl level because no plausible sources of additional extragalactic surface brightness can satisfy the multiple constraints that these sources ( 1 ) have not already been detected , ( 2 ) contribute total surface brightness comparable to the surface brightness contributed by detected galaxies , and ( 3 ) do not produce ebl fluctuations larger than the upper limits set by correlations in the object - masked hdf .
figure 3 shows that these constraints admit only extremely low surface brightness objects that must be so extended and numerous as to be confusion - limited on the sky .
the parameters of simple phenomenological models that meet these constraints would make such a population disjoint from the parameter space of all detected objects .
extrapolation of the hdf galaxy number counts to infinitely faint limits would add at most a few percent to the mean ebl .
figure 4 shows that a diffuse component that is as large as the surface brightness from detected galaxies would require a @xmath180-band angular clustering amplitude of @xmath181 , an order of magnitude less clustered than the faintest observed galaxies .
fluctuations in the ebl from intergalactic stars as seen in galaxy clusters could be as large as the upper limits on the fluctuations in the hdf , but would contribute only incrementally to the mean ebl .
in addition to constraining the plausible level of the optical ebl , these upper limits on small - scale sky fluctuations in the hdf provide a test for any proposed model for faint galaxy populations .
we plan to test galaxy population models that are more detailed than the simple phenomenological model described in section 4 ( paper iii ) .
we invite others to suggest possible models , compute the predicted autocorrelation function of sources below the detection limits , and compare with these limits from the hdf .
these correlations are also important for attempts to detect weak lensing signals in deep optical imaging ( van waerbeke et al . 1997 ; refregier et al . 1997 ) .
as we emphasize throughout this paper , the measured correlations of the object - masked sky in the hdf should be treated as upper limits on the true ebl fluctuations .
the wings of the wfpc2 point spread function almost certainly contribute to the measured fluctuations .
although there are reasons to exclude flat fielding errors or galactic light as the principal sources of the fluctuations , we can not rule out some contribution from these effects . to improve these upper limits
, we are examining these systematics and foregrounds in further detail ( paper ii ) .
the hdf provides the best constraint to date on small - scale ebl fluctuations because its superior resolution allows detection of faint galaxy populations that would be confusion - limited in ground - based imaging . in section 2 , we discuss how fluctuation analysis of shallower hst or ground - based images would be dominated on small scales by the profiles of galaxies that only the hdf detects .
however , ground - based imaging is required to study fluctuations on angular scales much larger than several arcseconds , where access to larger collecting area and accurate control of systematics outweigh the hst s advantages of high resolution and lower sky background .
if we measure object - masked sky correlations in deep ground - based images , we can now use the hdf galaxy counts and profiles to model and subtract the surface brightness correlations that are caused by galaxies that are fainter than the ground - based detection limits , but that were detected by the hdf .
fluctuation analysis of deep ground - based imaging will allow statistical tests for diffuse intergalactic light on larger angular scales and for low surface brightness companions to galaxies that were too bright to be included in the hdf ( see section 4c ) .
future hst observations will allow measurement of ebl fluctuations in independent fields and at fainter surface brightness levels than the hdf .
the planned hdf south ( october 1998 ) will reach to similar depth and allow us to test if the hdf north is typical .
the advanced camera for surveys ( scheduled for installation on hst in 1999 ) will have superior sensitivity and better - controlled systematic uncertainties than wfpc2 , as well as a @xmath182 contiguous field of view , thus allowing measurement of surface brightness fluctuations at fainter levels and somewhat larger scales .
the proposed gto program for acs focuses on deep imaging of galaxy clusters .
fluctuation analysis of the acs cluster fields will be an important test for sources of diffuse intergalactic surface brightness . in the @xmath183 m infrared ,
the proposed next generation space telescope might detect statistical fluctuations in the ebl from the first generation of stars ( e.g. , haiman & loeb 1997 ) even if such objects evade direct detection .
we thank r. williams and the entire hdf team , whose vision and effort made this project possible .
we acknowledge discussions with and many helpful comments from a. babul , r. bernstein , s. casertano , a. connolly , h. ferguson , a. fruchter , m. geller , e. groth , r. lupton , p. madau , j. peebles , a. refregier , p. stockman , m. strauss , a. szalay , t. tyson , and d. wilkinson .
we also thank r. bernstein and l. pozzetti for providing results in advance of publication .
this paper was initiated during a visit to the aspen center for physics .
support for this work was provided by nasa through grants hf-01078.01 - 94a and ar-05812.01 - 94a from the space telescope science institute , which is operated by aura , inc . under nasa contract nas5 - 26555 , williams , r. e. , blacker , b. , dickinson , m. , dixon , w. v. d. , ferguson , h. c. , fruchter , a. s. , giavalisco , m. , gilliland , r. l. , heyer , i. , katsanis , r. , levay , z. , lucas , r. a. , mcelroy , d. b. , petro , l. , postman , m. , adorf , h. m. , & hook , r. n. 1996 , , 112 , 1335 | statistical analysis of the unresolved light in the hubble deep field ( hdf ) strongly constrains possible sources of the optical extragalactic background light ( ebl ) .
this constraint is crucial for determining the spectrum of the ebl because reported upper limits on the optical ebl are several times larger than the surface brightness from detected galaxies , suggesting the possibility of additional galaxy populations . to test for the statistical signature of previously undetected sources
, we estimate the auto , cross , and color correlations of the `` sky '' in the hdf that remains after masking objects brighter than @xmath0mag .
auto and cross correlations of surface brightness in the @xmath1 and @xmath2 bandpasses are well - fitted by @xmath3 up to @xmath4 .
probable contributions of several instrumental systematics ensure that these correlations are firm upper limits on the true ebl fluctuations .
this measurement yields the most stringent limits to date on small - scale structure in the night sky ; analysis of shallower imaging would be dominated by galaxies now detected by the hdf . unless there is a truly uniform optical background ,
the mean ebl is likely to be within a small fraction of the surface brightness from detected galaxies .
no currently plausible sources of additional ebl satisfy the constraints that they ( 1 ) would not have already been detected , ( 2 ) contribute ebl comparable to that from detected galaxies , and ( 3 ) do not produce ebl fluctuations in excess of the upper limits set by correlations in the hdf .
these constraints admit only a confusion - limited population of extremely low surface brightness objects that is disjoint from the parameter space of all detected galaxies .
extrapolation of detected galaxy counts to zero flux would add only a few percent to the ebl .
diffuse intergalactic light clustered similarly to faint galaxies could explain some of the observed correlations but would contribute at most a few @xmath5 to the mean ebl .
450b_450 606v_606 i814i_814 # 1 |
the fiber bundle model , since its introduction @xcite , has been studied widely as a prototypical model of failure dynamics @xcite . this discrete element model , involving disorder and non - linear dynamics ( due to thresholds )
, also enables engineers to apply it to analyse the breaking properties of real materials ( e.g , , fiber reinforced composites @xcite ) and its simplicity and occasional analytic tractability @xcite has attracted statistical physicists , specifically for its intriguing dynamical critical behaviors @xcite .
there are mainly two extreme versions of this model .
both the versions consider a bunch of fibers ( hooke springs ) hanging from a rigid ceiling and a platform is connected to the ends of these fibers and a load hangs from that platform .
each fiber has a given limit or load - carrying threshold ( usually taken randomly from some distribution function ) , beyond which it fails .
completely different behaviors are observed when the elastic property of the lower platform changes .
two extreme cases arise when the platform is either absolutely rigid ( global load sharing case ) or absolutely soft ( though inextensible ; local load sharing case ) .
when a fiber breaks in the former case , its extra load is equally shared by all other remaining fibers , due to rigidity of the lower platform . in the latter case , however , due to local deformation of the platform , the load of the broken fiber is only to be carried by nearest surviving neighbours ( stress concentration occurs around the failure or breaking ) .
while in the global load sharing version the load at which the system completely fails scales with system size linearly , for the local load sharing case the increase is only sub - linear ( in fact @xmath0 ; @xmath1 being system size @xcite ) .
the implication being , the critical load per fiber @xmath2 at which the system fails , becomes finite for the global load sharing case and goes to zero in the large system size limit for the local load sharing case .
therefore the observations like divergence of relaxation time , proper scale free size distribution of avalanches , if any , are not seen in the local load sharing case ; unlike in the global load sharing version , where these can be analyzed in detail @xcite . ] .,width=264 ] however , the situation can be quite different if one makes the initial applied load localized ( at an arbitrarily chosen central site ; see fig .
[ setup ] ) in a local load sharing fiber bundle model in greater than one dimension ( for one dimension the damage interface can not increase ) .
let this load be increased at a slow but constant rate .
initially no load is present on any fiber except for the one at the central site .
as the applied load increases beyond the failure threshold of this central fiber , it breaks and the load carried by it is redistributed among its nearest neighbours and so on . here
we study two versions of the model . in model
i : in general , whenever a fiber breaks , the load carried by that fiber is redistributed equally among its nearest surviving neighbour(s ) . in this way , the fibers which are newly exposed to the load , say , after an avalanche , have a relatively low load compared to the ones which are accumulating load shares from the earlier failures and are still surviving .
as we shall see later , this helps in maintaining a compact structure of the cluster or patch of the broken fibers .
this local force redistribution is justified from the point of view that the newly exposed fibers are presumably further away from the point of loading and therefore have to carry a smaller fraction of the load at the original central site . the fibers on the perimeter of the failed or damaged region , which together are carrying the entire load , increase in number with time .
hence the load per fiber decreases during an ongoing avalanche , which is assumed to be a much faster process compared to the external load increase .
however , as the load on the bundle increases at a constant ( but slow ) rate , the load per fiber along the boundary will tend to increase .
eventually , a dynamically stable state will occur when the load per fiber will fluctuate around a stable value and the system has reached a self - organized state .
in this dynamical state , failure of fibers in the process of avalanches is seen to have a scale free size distribution , suggesting the state to be a self - organized critical one .
we study this model numerically to estimate the avalanche size and other exponent values .
these values are close to those found for stochastic sandpile model or manna model , within our numerical accuracies .
we then study a simpler version of this model , model ii , in which the load of a broken fiber is equally shared by all the surviving fibers that have atleast one broken neighbour , i.e. it is equally distributed along the boundary of the broken or damaged patch . due to the fact that local fluctuations are ignored in the process of load redistribution , this version is analytically tractable using a mean - field like approach and the numerical results compare well . as can be seen from fig .
[ config1 ] , this version of the fiber bundle model ensures an advancing interfacial ( mode - i ) fracture . this topic
is widely studied over decades both theoretically and experimentally ( see @xcite for a recent review ) .
particularly , in the plexiglas experiment @xcite two plates were taken and disorder was introduced by sandblasting and then were joined together , making a transparent block with an easy plane .
interfacial ( mode - i ) fracture was then studied with it .
a similar situation with this model would arise if one of the plates ( or softer surface ) could be pulled from the middle .
the advantage of the present situation for experiment and in particular for simulations is that one is free from the need of introducing a cut - off scale by hand as dissipation comes naturally in this model with the increase of effective system size with dynamics .
this is a more desireable situation since artificial dissipation scales often cause problems in estimating the exponent values , which had to avoided by measuring a different quantity instead @xcite .
below we first present the numerical results for the nearest neighbour load sharing model ( model i ) and then go over to the simpler version of uniform load sharing along boundary ( model ii ) , giving the analytical estimates for the later and comparing them with numerical results .
as mentioned before , usually one does not find a finite critical load for local load sharing fiber bundle models .
this is the result of extreme value of statistics , in which the strength of the sample is determined solely by its weakest point . in that case , as the load , no matter how small , is applied system wide , there will always be a large enough weak patch , which will be broken and due to concentration of the load on the boundary and the fact that more load is nucleated if the fracture progresses , the patch keeps growing , leading to a system wide failure without any further input from outside .
however , the situation will be very different when the initial load is applied locally and not on all elements of the system .
initially we apply a load on one fiber , say the one at the middle .
we then continue to increase the load at a very slow rate .
when this fiber breaks , the load carried by it is redistributed equally among its four neighbours . since we keep on increasing the load ( only on the fibers that are already carrying a non - zero load ) , this breaking and redistribution dynamics continues . in general , whenever a fiber breaks , the load carried by that fiber is equally shared by the surviving fiber(s ) nearest to it .
when the islands break ( all four neighbours already broken ) , its load is redistributed among its nearest surviving neighbor(s ) ( searched along the perpendicular axes ; other searchings are slow and do not change the critical behavior ) , no matter how far they are located . note that in this way of redistribution , the fibers that are newly exposed to the load have a rather low load per fiber compared to the ones which are gathering and withstanding the loads of broken fibers from a few steps earlier .
this ensures that the compact structure of the patch of the broken fibers is maintained , since a rather weak but new fiber is more likely to survive than a stronger but old one , thereby eliminating the possibilities of fingering like effects from the patch boundary .
also note that , following the usual definition of avalanche , we do not increase the external load when a fiber breaks or more fibers break due to load redistribution .
in other words , the process of load redistribution happens in a much faster time scale than load increase , which is quasistatic . , i.e. the number of fibers sharing the load is plotted against time for both models for a single realisation .
it increases on average but has jumps of increase and decrease as well .
( b ) and ( c ) show the probability distribution @xmath3 of the sizes of the jumps @xmath4 for model i and model ii respectively , all measured after the time the system has reached a steady state.,width=340 ] ] ) for nearest surviving neighbour load redistribution ( model i ) .
the saturation is close to @xmath5 .
the inset shows the stationary distribution of the load per fiber value ( @xmath6 ) and the failure threshold values ( @xmath7 ) of the surviving fibers on the boundary.,width=340 ] as we shall see , the application of load at a point and subsequent localized redistribution rules , bring the system into a self - organized critical state .
one of the signatures of self - organization is the stationarity in macroscopic quantities . in this case
, the most important quantity is the average value of the load per fiber @xmath8 where @xmath9 is the slowly increasing load and @xmath10 is the effective system size , i.e. number of fibers carrying the load , which also increases with time ( see fig .
[ jump ] ) .
we find that this quantity stabilises , i.e. becomes time independent after some initial transients .
the load carried by each fiber is not same .
we measured the probability density function @xmath6 of the load of the fiber , i.e. the probability that a fiber has load between @xmath11 and @xmath12 is @xmath13 .
in fact the entire distribution of the load per fiber , and not only its first moment , becomes stable .
similarly , we also find stability in the probability density function of the failure threshold values of the surviving fibers @xmath7 , each of which carry a non - zero load .
again , @xmath14 denotes the probability that a surviving fiber carrying a non - zero load has the failure threshold value between @xmath11 and @xmath12 . in fig .
[ load - sat - loc ] the saturation of the average load per fiber value is shown .
the inset shows the stationary distribution of the load per fiber values ( @xmath6 ) and failure threshold values ( @xmath7 ) . while these results are for the initially uniform ( in [ @xmath15 ) threshold distribution , the same phenomenon occurs for other threshold distributions , like weibull , gaussian , triangular etc .
the average stationary values change for different distributions .
later we will give analytical estimations of these stationary distributions for a simplified model , which match well with simulation results .
note that usually the two basic ingredients of a self - organized critical system are external drive and dissipation @xcite . in the present case , although we have an external drive , there is no explicit dissipation .
but the effect of dissipation enters the system from the fact that the effective system size ( @xmath10 : number of surviving fibers carrying non - zero load ) is an increasing function of time .
as indicated before , we follow a quasistatic load increase protocol . in effect , this means increasing the load per fiber uniformly on all those surviving fibers that already carry a non - zero load , until the weakest ( having the smallest difference between load and failure threshold ) fiber breaks .
then we wait for the system to adjust by breaking the fibers and redistribution of the loads , and come to a state when no further fiber breaks .
all the fibers broken in between , constitute one avalanche .
this process is repeated in every time step .
the size distribution of avalanches @xmath16
( @xmath17 being the size of avalanche , i.e. number of fibers broken in an avalanche ) measured in this way are plotted in fig .
[ avalanche - loc - comb ] .
this shows a power - law distribution @xmath18 with exponent value close to @xmath19 .
this exponent value remains unchanged when other types of threshold distributions ( not shown here ) are used .
we have also measured the duration of an avalanche .
this is defined as the number of times ( @xmath20 ) the load is redistributed during one process of avalanche .
the probability distribution of duration @xmath21 is also a power - law decay ( @xmath22 ) with exponent value close to @xmath23 ( see inset of fig . [ avalanche - loc - comb ] ) .
it is interesting to note that avalanche size distribution exponent value is rather close to the stochastic sandpile or manna universality class @xcite .
apart from these , the size distributions of the changes in effective system size ( @xmath4 ) was measured , again after the system has reached a stationary state .
this distribution is also a power - law with exponent value close to @xmath24 ( see fig .
[ jump ] ) . .
inset : distribution of the duration of an avalanche is shown , which is also a power - law decay with exponent value @xmath25.,width=340 ]
a simpler version of the model studied so far is the one where the load carried by a broken fiber is equally redistributed among all the fibers that have at least one broken neighbour .
this is another way of saying that the load is redistributed uniformly along the entire boundary , once a fiber breaks ( see fig .
[ config2 ] ) . ] .,width=264 ] since the size of the boundary is large after the transients , the behaviour of the model in the present form is tractable via a mean field like approach .
below we first discuss the stationary distribution functions of load per fiber value and failure thresholds for different initial failure threshold distributions ; comparing the estimates with numerical results .
then we do the same for avalanche size distribution . ] ) for model ii ( i.e. , uniform load redistribution along the patch boundary ) .
the solid line is the analytical estimate ( @xmath26 ) and the simulation result is close to that value .
the inset shows the comparison of the simulation results and analytical estimates of the saturation load per fiber value , when the threshold is distributed uniformly between [ @xmath27.,width=340 ] same as the earlier version of the model , the load distribution and the threshold distribution along the patch boundary reaches a stationary state after some transient . in this case , however , one can also make an analytical estimate along with the numerical simulations , to get a good agreement .
consider the failure thresholds to be distributed uniformly between the limits [ @xmath27 .
first let us take the case when @xmath28 .
we follow the algorithm for load increase as stated before .
if one measures the average load per fiber with time @xmath29 ( one time step being one avalanche ) , one finds that after a transient it saturates to a given value ( see fig .
[ load - sat ] ) . for this case , the value is close to @xmath26 . in analytically estimating it we assume that dynamically all values of force ( between [ @xmath15 , i.e. the range of the threshold distribution ) are generated with equal probability .
so , the probability that a load between @xmath11 and @xmath12 is applied on the fibers is @xmath30 ( unnormalized ; @xmath31 is a constant independent of @xmath11 since the force distribution is assumed to be uniform ) . now among these fibers , only the ones with failure threshold greater than @xmath11 will survive .
the probability that the fibers have failure threshold greater than @xmath11 is proportional to @xmath32 ( since we are considering threshold distribution to be unuiform in [ @xmath15 ) .
therefore , the probability that a surviving fiber has a load between @xmath11 and @xmath12 is the joint probability that a load between @xmath11 and @xmath12 is applied on the fibers _ and _ the fibers have failure thresholds higher than @xmath11 , which is given by @xmath33 . from normalization @xmath34 , one has @xmath35 .
so , the probability density function of the surviving fibers having load @xmath11 is @xmath36 , which compares well with numerical estimate ( fig .
[ dist ] ) . of course
, the average load per fiber is @xmath37 , which is again close to what we see numerically ( fig . [ load - sat ] ) .
similarly , the probability that the fibers in the system have a failure threshold between @xmath11 and @xmath12 is @xmath38 ( unnormalized ; @xmath39 is independent of @xmath11 since the threshold distribution is uniform in [ @xmath15 ) .
now , the probability that a load equal to or less than @xmath11 is applied on the fibers is proportional to @xmath11 ( since force distribution is uniform ) .
hence the probability that a surviving fiber has a threshold between @xmath11 and @xmath12 is the joint probability of a fiber having its threshold between @xmath11 and @xmath12 _ and _ load below @xmath11 .
this probability is @xmath40 .
normalization gives @xmath41 i.e. , @xmath42 .
so , the probability density function for the threshold to be @xmath11 is @xmath43 .
this also compares well with simulations in fig .
[ dist ] .
now we go to the case when @xmath44 , i.e. , the threshold distribution function has a lower cut - off .
now , since all fibers have threshold higher than @xmath45 and we have assumed before that force is uniformly distributed , the probability density function for force will be uniform between [ @xmath46 and then decreasing linearly as before ( can be obtained along the same line as before ) . from normalization condition ,
the height of the uniform part is @xmath47 .
therefore , one gets @xmath48\quad \mbox{otherwise}\end{aligned}\ ] ] this is what we get numerically as well ( see fig .
[ dist ] ) for different values of @xmath45 ( for @xmath28 we get back @xmath49 ) . in the same line
one can also calculate the probability density function for threshold with finite @xmath45 @xmath50 this compares well with numerical result ( also shown in fig .
[ dist ] ) .
if we now calculate the average force , it comes out to be @xmath51 ,
\label{avgsigma}\ ] ] we compare the prediction of this calculation with the values obtained numerically in fig .
[ load - sat ] . of course
, we get back @xmath26 for @xmath28 . to show universality of the above phenomenon with respect to different threshold distributions , we have checked it for different threshold distributions ( gaussian , triangular etc .
we show the case of weibull distribution , which is more abundant in real situations . ) and failure thresholds ( @xmath7 ) of the fibers in the boundary of the broken patch for weibull threshold distribution with @xmath52 and @xmath53 for model ii .
the solid lines are analytical predictions ( eqs .
( [ wei - force ] ) and ( [ wei - th ] ) ) and the points are simulation results . the inset shows the time variation of the average load per fiber value , which saturates close to the analytical estimate @xmath54.,width=340 ] a general form of the weibull distribution is @xmath55 where @xmath56 and @xmath57 are two parameters .
let us consider the particular case when @xmath52 and @xmath53 .
as before , the probability that the threshold is greater than @xmath11 is proportional to @xmath58 .
now since the probability density function for force is uniform , the probability of a fiber having load between @xmath11 and @xmath12 is @xmath59 , with @xmath60 ( unnormalized ) since the force distribution is assumed to be uniform . as before , the normalization condition would require @xmath61 giving @xmath62 .
hence the normalized probability density function for the load on the surviving fibers will be @xmath63 similarly , the probability that the load is lower than @xmath11 is proportional to @xmath11 . using the form for threshold distribution ( @xmath64 ) , it is straightforward to get the probability density function for threshold distribution of the survived fibers , which is @xmath65 both of these functions are in good agreement with numerical simulations ( see fig . [ weibull ] ) .
also the saturation value of the average load per fiber can be calculated as @xmath66 which is also in good agreement with simulation value ( see inset fig . [ weibull ] ) .
the size distribution of avalanches are measured as before .
now we find the size distribution exponent to be close to @xmath67 ( see fig .
[ avalanche - comb ] ) , which is in agreement with the scaling prediction of avalanche size distributions in soc models for mean field @xcite , i.e. @xmath68 , where @xmath69 and @xmath70 are the surviving probability decay exponent and number of particle growth exponent respectively , both of which are 1 in mean field , giving @xmath71 .
note that similar exponent value was obtained before when one measures avalanches only near the critical point in global load sharing models @xcite and also in other versions of self - organized models with fiber regeneration @xcite .
we also measure the distribution of avalanche duration .
the duration distribution is a power - law with exponent value close to @xmath72 ( see inset of fig .
[ avalanche - comb ] ) , which is again in agreement with scaling predictions of soc models in mean field , i.e. @xmath73 .
the size distribution of the changes in effective system size is also close to @xmath74 ( see fig . [ jump ] ) . , which is also our estimate from scaling arguments .
inset : the distribution of the duration of an avalanche is plotted for model ii .
this shows a power - law decay with exponent value @xmath72.,width=340 ] now to estimate the value of the avalanche size exponent , we first assume that the average load per fiber on the perimeter of the damaged region has a distribution which is gaussian around its mean : @xmath75 .
hence , on a dimensional analysis , mean - squared fluctuation @xmath76 . also the avalanche size @xmath17 scales as @xmath77 since it may be viewed as the number of broken fibers after a load increase of @xmath78 .
this gives @xmath79 the probability of an avalanche being of the size between @xmath17 and @xmath80 is @xmath81 .
now , the deviation from the critical point scales @xcite with the cumulative size of all avalanches upto that point ; giving @xmath82 .
if we take @xmath18 , then @xmath83 comparing eqs .
( [ sig1 ] ) and ( [ sig2 ] ) therefore we have @xmath84 .
so the probability density function for the avalanche size becomes @xmath85 , which fits well with simulation results ( fig .
[ avalanche - comb ] ) . note
that we have also checked these results for triangular lattices , where the nearest neighbours are six .
as one can see , the analytical estimates are independent of lattice topology , hence predict same behaviour for triangular lattice as well , which is what we get numerically as well ( not shown here ) .
here we study the local load sharing fiber bundle model in a square lattice , with the load being initially applied only at a central site .
the load is slowly increased and a broken patch grows in the middle of the system , the boundary of which carries the entire load . since the boundary increases , the load per fiber reaches a stationary value after some time and responses of the system in that state is scale free , indicating a self - organized critical state for the system .
we study two versions of this model . in model
i : after the failure of the central fiber , the load is shared by its four nearest neighbouring sites , which initially had no load on them .
once any of them fails , its load will be shared equally by its nearest surviving fiber(s ) and so on .
these load shares accumulate on all the fibers on the perimeter and grows with time since their respective appearance on the perimeter of the damaged central region .
this may be interpreted as the fact that further the surviving fibers are from the central point , lower are the loads on them and closest fiber ( also the longest surviving fiber on the perimeter ) to the central site accumulates the maximum load share . as stated before
, the model reaches a stationary state after some transients .
this leads to a self - organized dynamics of failure in the model .
we have studied the breaking statistics in such a self - organized dynamical state .
while the usual ingredients for self organization are external drive and dissipation @xcite , in our model the latter is apparently absent .
however a similar effect is coming from the fact that the effective system size , i.e. the perimeter length of the patch of the broken fibers , is not a constant but increases with the increase of load with time .
this leads the system into a self - organized critical state , from which if one makes a departure ( by increasing the load ) the response of the system ( size of the avalanche and its duration ) can be of any size , leading to scale free dynamics .
particularly , we observe , in the avalanche size distribution , exponent value is @xmath19 ( see fig .
[ avalanche - loc - comb ] ) and the duration distribution of an avalanches has an exponent value @xmath25 ( see inset of fig .
[ avalanche - loc - comb ] ) .
also the change in the effective system size follows a power - law with exponent close to @xmath86 ( see fig . [ jump ] ) .
we have checked , the exponent values do not change if the identities of the nearest neighbour fibers ( along lattice axes , or along diagonals as well ) are changed .
also we have checked the universality using different ( e.g. , weibull , uniform between [ @xmath27 and so on ) fiber strength distributions .
these exponent values are close to what are seen in manna model .
we then studied a simpler version of this model , model ii , where the load redistribution is uniform along the boundary of the broken patch .
this reduces the fluctuation in the system significantly ( see figs .
[ config1],[config2 ] for a visual comparison ) , and brings the system to a mean - field limit .
the average load per fiber value still saturates to a finite value .
we have made an estimate of this average value for a system with uniform threshold distribution between [ @xmath27 , and also for a weibull distribution with certain values of parameters .
it turns out , for uniform distribution with @xmath28 , the probability density function for a surviving fiber to have load @xmath11 can be calculated to be @xmath49 and average value of @xmath87 is estimated to be @xmath26 , which is in close agreement with simulation results .
however , for non - zero values of @xmath45 , we find significant departure from the analytical estimate ( given by eq .
( [ avgsigma ] ) ) .
similar estimates for weibull distribution with @xmath52 and @xmath53 is made , where this saturation value comes out to be @xmath54 , again very close to simulation results .
we also present a scaling theory to predict the avalanche size distribution in this case .
the exponent value for the size distribution is estimated to be @xmath67 , which is in close agreement with the numerical results ( fig .
[ avalanche - comb ] ) .
this is of course the mean field limit of self - organized criticality @xcite .
the exponent value ( @xmath88 ; see inset of fig .
[ avalanche - comb ] ) for the distribution of duration of avalanche also matches with the mean field estimate .
note that one could not have obtained this self - organization in an one dimensional array of local load sharing fibers , due to the finite extension of the damaged region ( two fibers on the two sides of the damaged region will have to carry the load ) .
this self - organization therefore occurs only for two and higher dimensions .
also note that a long range load transfer , with initial load along a line , requires an explicit dissipation term to get self - organized dynamics @xcite .
similar efforts were made in capturing the morphology of the fractured surfaces by considering the elastic modulus of the lower plane of the model @xcite . in conclusion ,
the self - organizing dynamics of failure in this local load sharing fiber bundle model ( in more than one dimension ) has interesting critical ( scale free ) behaviour .
the models show mode - i damage propagation without having to put a dissipation scale by hand . in the nearest neighbour load sharing version ( model i ) , the avalanche size distribution follows a power - law with exponent value close to @xmath19 .
the distribution for the duration of avalanche follows a power - law with exponent value close to @xmath25 .
we have checked that these responses remain unchanged with different types of distribution of the breaking thresholds of the fibers and both square and triangular lattices .
we have also studied a simpler version of this model , where the load redistribution is uniform along broken patch boundary ( model ii ) , where mean - field approximations work and find good agreement between analytical predictions and numerical simulations .
the authors acknowledge the participants of the indnor project meeting , particularly alex hansen , knut mly , srutarshi pradhan and purusattam ray , for useful discussions .
the financial supports from project indnor 217413/e20 ( nfr , govt . of norway ) and bkc s jc bose fellowship ( dst , govt . of india ) are acknowledged . , edited by h. j. herrmann and s. roux ( north - holland , amsterdam , 1990 ) ; b. k. chakrabarti and l. g. benguigui , _ statistical phtsics of fracture and breakdown in disordered systems _ ( oxford university press , oxford , 1997 ) ; m. sahimi , _ heterogeneous materials ii : nonlinear and breakdown properties _ ( springer - verlag , new york , 2003 ) .
f. kun , f. raischel , r. c. hidalgo and h. j. herrmann in _ modelling critical and catastrophic phenomena in geoscience : a statistical physics approach _ , edited by p. bhattacharyya and b. k. chakrabarti ( springer - verlag , berlin ) , vol .
705 , p. 57 | we study the dynamics of a local load sharing fiber bundle model in two dimensions , under an external load ( which increases with time at a fixed slow rate ) applied at a single point . due to the local load sharing nature , the redistributed load remains localized along the boundary of the broken patch .
the system then goes to a self - organized state with a stationary average value of load per fiber along the ( increasing ) boundary of the broken patch ( damaged region ) and a scale free distribution of avalanche sizes and other related quantities are observed . in particular , when the load redistribution is only among nearest surviving fiber(s ) , the numerical estimates of the exponent values are comparable with those of the manna model . when the load redistribution is uniform along the patch boundary , the model shows a simple mean - field limit of this self - organizing critical behaviour , for which we give analytical estimates of the saturation load per fiber values and avalanche size distribution exponent .
these are in good agreement with numerical simulation results . |
the search for high - redshift galaxies and the effort to map the star formation history of galaxies have progressed rapidly in the last several years as magnitude limited spectroscopic surveys pushed into the @xmath22 range ( cowie et al .
1996 ; cohen et al . 1996 ) , while color based selection techniques produced many objects in the @xmath23 range ( steidel et al .
1996a , 1996b ; franx et al .
1997 ; steidel et al .
1998 ) , particularly in the exquisite hubble deep field ( hdf ) sample ( lowenthal et al . 1997 ) . however , the galaxies chosen by these techniques correspond to objects with ongoing massive star formation and small amounts of extinction , and may represent only part of the populations at these early epochs .
more evolved objects may be heavily dust reddened and more easily picked out at submillimeter wavelengths ( smail , ivison , & blaine 1997 ; hauser et al .
1998 ) , while earlier stages in evolution may have relatively little continuum light and be too faint to be seen in the magnitude limited samples or selected with the color - break techniques at current sensitivity limits .
this latter class of objects may represent the earliest stages of the galaxy formation process , in which substantial amounts of metals have yet to form .
these galaxies may have much stronger ly@xmath16 emission relative to the stellar continuum , since they have massive star formation that can excite the ly@xmath16 emission line , but without so much dust that the line is suppressed , and this can result in very high observed equivalent widths in the range of 100200(1+@xmath11 ) ( e.g. , charlot & fall 1993 ) .
such objects may be hard to pick out with color - break techniques but be detectable in ly@xmath16 searches of sufficient depth ( cowie 1988 ; thommes 1996 ) .
an increased incidence of strong ly@xmath16 emission does , indeed , appear in the color - break samples at fainter continuum magnitudes ( steidel et al .
1998 ) . earlier blank - field ly@xmath16 surveys ( e.g. , thompson , djorgovski , & trauger 1995 ; thompson & djorgovski 1995 ) failed to find such objects , as pritchet ( 1994 ) has summarized , but their sensitivity lay at the margin of where such objects would be expected in significant numbers ( cowie 1988 ; thommes & meisenheimer 1995 ) .
however , hu & mcmahon ( 1996 ) , using very deep targeted narrowband searches ( @xmath24 ergs @xmath8 s@xmath9 ) , found @xmath25 ly@xmath16-emitting galaxies with the very strong emission and weak or undetected continuua predicted for early star - forming objects .
the advent of 10 m telescopes , along with this successful detection of ly@xmath16 emitters , prompted us to undertake a new survey that has been successful in detecting blank - field high-@xmath11 ly@xmath16 emitters .
the present letter describes the early results of this search , which uses extremely deep narrowband filter exposures taken with lris ( oke et al .
1995 ) on the keck ii telescope to search for emission - line populations at extremely faint flux levels ( @xmath26 ergs @xmath8 s@xmath9 ) .
this survey picks out sources of extremely high equivalent width ( @xmath27 ) emission lines as candidates , and then uses followup lris spectroscopy ( 2 ) to determine if these are ly@xmath16 emitters .
the first results from the hawaii survey with a 5390/77 filter , corresponding to ly@xmath16 emission at @xmath5 , were described in cowie & hu 1998 ( hereafter , paper i ) , and yielded a number of candidate high-@xmath11 galaxies similar to the redshift @xmath25 ly@xmath16-emitting galaxies found by hu & mcmahon ( 1996 ) and emitters at @xmath28 ( pascarelle et al .
1996 ; francis et al .
1997 ) in targeted searches . in paper
i , we showed that the use of color selection on emission - line selected objects of high equivalent width ( @xmath29 ) picked out objects with continuum colors similar to those of color selected lyman break galaxies with measured redshifts of @xmath5 .
they also recovered the one field object whose previously measured redshift placed ly@xmath16 within the filter bandpass .
however , the emission - line galaxies selected by their high equivalent width also comprised objects with very faint continuua , that would have fallen below the magnitude threshold of current lyman break surveys , and also included two objects that could not be detected in keck imaging of the optical continuum ( @xmath30 @xmath31 , @xmath32 , and @xmath33 ; @xmath34 ) . in the present letter we first present spectroscopic followups for the narrowband candidates of paper i. in a small fraction of the cases , the spectrum shows both ly@xmath16 and c @xmath35 1550 , confirming the redshift identification but suggesting agn - like properties
however , the majority of the spectra show only a single strong emission line .
the absence of other detectable features , in combination with the high equivalent width of the selected candidates , identifes the single line as redshifted ly@xmath16 emission , and argues that the equivalent width criterion ( @xmath36 ) can , in fact , be used as a good diagnostic of high-@xmath11 ly@xmath16-emitting galaxies .
we then ( 3 ) present results of a second deep narrowband search with a 78 bandpass @xmath37 filter ( ly@xmath16 at @xmath38 ) and followup spectroscopy , that confirmed two ly@xmath16-emitting galaxies at this higher redshift . in 4 , we describe a very deep blank - field spectroscopic search ( 6 hr lris integration covering @xmath39 ) that yielded four emitters at redshifts 3.04 5.64 .
finally ( 5 ) , the data on emission - line objects from the imaging surveys in the two redshift intervals are combined with various spectroscopic surveys at lower redshift to show the evolution of the emission - line fluxes with redshift .
we argue that the ly@xmath16-emitting objects are significant contributors to the integrated star formation of the galaxy population throughout the @xmath14 redshift range , and that the integrated star formation rate of the ly@xmath16 selected objects is flat , or possibly rising through this redshift range , with a value greater than 0.01 @xmath40 mpc@xmath18 for @xmath20 and @xmath41 . at the highest redshifts ,
most of the star formation may be occurring in objects of this class .
paper i provides a sample of 12 objects in the ssa22 and hdf fields , with @xmath42 in the 5390/77 filter .
a further 7 objects were detected in a similar field surrounding the @xmath43 quasar br00191522 ( storrie - lombardi et al .
1996 ) , giving a total sample of 19 objects over an area of @xmath44 .
all the ssa22 and hdf candidate objects have now been spectroscopically observed using lris on the keck ii telescope , as have three of the objects in the br00191522 field .
multi - slit masks with @xmath45 slits were used , giving a resolution of 16 ( cowie et al .
1996 ) , and exposure times ranged from @xmath46 hrs , with most objects having the longer exposure .
two objects in the hdf field have strong c emission , placing them at @xmath47 but suggesting agn activity ( fig . [ fig:1 ] ) .
these objects appear similar to the brighter objects in pascarelle et al .
. however , the remaining objects have only one strong emission line , and the continuua are too weak for absorption line identifications . figure 1 .
spectra of strong ly@xmath16 emitters in ssa22 , la1 and la3 , ( _ top two panels _ ) compared with the spectrum of an [ o ] emitter with extremely high equivalent width ( _ third panel _ )
shifted so the [ o ] emission lies at the narrowband wavelength ( 5390 ) .
we also show for comparison one of the two ly@xmath16 emitters ( la1 ) in the hdf field that shows strong c emission ( _ bottom panel _ ) .
the observed equivalent widths of the two ssa22 objects are @xmath48 for the topmost object and @xmath49 for the second .
the observed equivalent width of the [ o ] line in the comparison object is 220 .
the h@xmath50+[o ] complex is clearly evident in the case of the [ o ] emitter , and these cases may be easily distinguished from genuine high-@xmath11 ly@xmath16 emission objects.[fig:1 ] a primary concern in the simple equivalent width selection procedure is whether we are misinterpreting emission lines such as [ o ] 3727 and h@xmath16 6563 , that are common features in galaxies at moderate redshift , as higher redshift ly@xmath16 emission . for objects
where the continuum is strongly detected it is possible to use color information to diagnose the different lines ( paper i ) , but in cases in which only one emission line is seen , we can only make the distinction based on the equivalent width strength of the line , assuming that an observed equivalent width @xmath51 is generally ly@xmath16 rather than unusually strong [ o ] or h@xmath16 .
spectroscopic followup is therefore essential to decide whether this procedure is reliable .
it is easiest to argue that single - line objects are ly@xmath16 emitters at lower redshifts , where the number of possible contaminating lines is smaller .
an emission line at 5390 lies below rest - frame h@xmath16 , and only slightly above the [ o]+h@xmath50 complex at 5000 , so that the only probable sources of confusion are [ o ] 3727 emitters at @xmath52 .
however , such objects would then have h@xmath50 4861 and [ o ] 5007 lying at 7030 and 7241 , and we would expect these emission lines to be strong , given the necessary strength of the [ o ] 3727 line . in figure [ fig:1 ]
we compare the spectra of two strong emitters ( @xmath53 and @xmath49 ) found in the ssa22 field with one of the strongest equivalent width [ o ] emitters ( @xmath54 ) found in the hawaii deep field survey of cowie et al .
( 1996 ) , and with one of the objects with ly@xmath16 and c emission detected in the hdf .
the [ o ] comparison spectrum has been wavelength shifted so that its [ o ] line lies over the emission lines in the objects .
it can be clearly seen that the absence of the h@xmath50+[o ] complex in the ssa22 spectra rules out the possibility that the objects are [ o ] emitters , and leaves us only with the ly@xmath16 interpretation . 0.25 in figure 4 .
the plots show ly@xmath16 emission - line objects recovered from deep multi - slit spectroscopy of a field around the quasar br12020725 using lris .
the 4 detected objects have observed equivalent widths of @xmath55 , @xmath55 , 450 , and 350 .
see fig . 3 for the corresponding two - dimensional spectra .
if the emission line were [ o ] , the h@xmath50+[o ] complex would lie in the @xmath56 region.[fig:4 ] at higher redshifts , the most problematic contaminants are likely to be high ionization extragalactic h regions , with strong lines of [ o ] and h@xmath16 .
spectroscopic followups of candidate emitters from narrowband surveys where the filter bandpass lies in regions devoid of strong night sky lines , as is the case for the 5390 and 6741 filters , can rule out [ o ] by the doublet line structure .
h@xmath16 from high ionization systems can have observed equivalent widths of several hundred angstroms ( cf . , stockton & ridgway 1998 ) , and requires broad wavelength coverage of the [ o ] lines and/or the measurement of a continuum break across the emission line to confirm it as ly@xmath16 .
we do not find such contaminants evident among the narrowband and spectroscopic samples discussed in the present paper .
narrowband observations of the ssa22 field were obtained on the nights of ut 1997 september 2122 with the lris camera on keck ii , through a narrowband interference filter centered at 6741 , with a bandpass of 78 and transmission of 81% .
the net exposure of 2.7 hrs was obtained as a sequence of 900 s exposures taken with @xmath58 offsets , with fwhms of @xmath59 ( 1.2 hrs ) and @xmath60 ( 1.5 hrs ) .
a corresponding 1.2 hr @xmath61-band image was obtained as a series of 360 s exposures to define the local continuum .
three @xmath42 objects were detected , which are marked on the narrowband image in figure [ fig:2 ] ( plate 1 ) , and shown as 2-dimensional spectra in figure [ fig:3 ] ( plate 2 ) . at this narrowband wavelength ( 6741 )
we may be contaminated by [ o ] 3727 at @xmath62 , h@xmath16 at @xmath63 , or [ o ] or h@xmath50 at @xmath64 .
however , our spectroscopic followup again easily distinguishes such cases .
the lowest flux object , with the weakest equivalent width constraint ( @xmath65 ) , is identified as an [ o ] emitter at @xmath66 , but the stronger object with substantial red wavelength coverage does not have detected [ o ] and does not permit this interpretation . for the second strong object there is a strong break across the emission line , consistent with the ly@xmath16 interpretation . nor does either of the objects have emission lines near the primary line , that would permit the objects to be identified as [ o ] or h@xmath50 .
once again there appears no choice but to interpret these lines as ly@xmath16 . 0.4 in figure 5 .
composite plot of h@xmath16 flux times redshift squared versus redshift . at low redshift
we show the @xmath67+[n])@xmath68 fluxes ( _ filled squares _ ) from the @xmath69 sample of gallego et al .
( 1995 ) , converted to corresponding h@xmath16 flux by multiplying by 0.75 .
the crosses show normalized h@xmath16 fluxes or converted , normalized [ o ] fluxes from the hawaii wide - field surveys ( songaila et al .
the filled triangles show renormalized , converted [ o ] fluxes from a @xmath70 spectroscopic sample ( cf .
, cowie et al .
1997 ) with open squares showing renormalized @xmath67+[n])@xmath68 or with circles showing renormalized @xmath71(ly@xmath16 ) at the highest redshifts if the spectrum does not cover [ o ] .
finally , the remaining circles ( at @xmath72 and @xmath73 show @xmath74 for the samples of objects detected in the two narrowband filters .
the solid ( @xmath20 ) and dashed ( @xmath75 ) reference lines show the fluxes corresponding to star formation rates @xmath76 = 10 , 1 , and 0.1 @xmath40 yr@xmath9 for @xmath77 km s@xmath9 mpc@xmath9 and the @xmath76 normalization discussed in the text.[fig:5 ]
the surface density of emitters from the narrowband search is sufficiently high ( @xmath78/@xmath10/unit @xmath11 at @xmath79 , based on the 3 fields ) that searches of existing very deep long - slit or multi - slit lris observations should yield such objects over the optical redshift range ( @xmath80 ) .
indeed , franx et al .
1997 identified a second @xmath81 object by inferring that the single strong emission line seen in a serendipitously placed slit during spectrosopic exposures of their color selected , gravitationally lensed galaxy , was also due to ly@xmath16 at their target object s redshift . in our own archival lris data
the best suited exposure is that around the quasar br12020725 , where a 6 hr integration in a single mask configuration was obtained under photometric conditions and with reasonable seeing ( ut 31 march 2 april 1997 ; 4 hrs obtained with @xmath82 fwhm , and the remaining 2 hrs with @xmath83 fwhm ) . the data were taken to study color selected and emission - line selected objects in the quasar field ( hu , mcmahon , & egami 1996 ; giallongo et al .
1998 ) and the mask comprised only twelve slits , with the blank sky area covering roughly 200 arcsec@xmath84 .
a search of this spectral image yielded four extremely high equivalent width emitters ranging in redshift from @xmath85 .
these spectra are shown in fig .
[ fig:4 ] and in plate 2 .
the two highest redshift objects can be seen in the red continuum , and have a strong break across the emission line , consistent with the expected ly@xmath16 forest decrement .
the surface density of emitters in the spectroscopic field is roughly 60,000 deg@xmath86/unit @xmath11 in the accessible redshift range , consistent with the narrowband searches when account is made for the slightly fainter flux limit in the spectrocopic data ( @xmath12 ergs @xmath8 s@xmath9 ) and the current uncertainty in the numbers .
ultimately , the census of the populations is best done with the narrowband data with its uniform selection procedures and well - defined flux limits , but the present data suggest that we will see substantial populations of these objects to very high redshift .
since resonant scattering enhances the effects of extinction , it is harder to convert the ly@xmath16 emission into a massive star formation rate than it is for line luminosity diagnostics such as h@xmath16 and [ o ] 3727 . for the present calculation , we assume that extinction may be neglected in computing the required massive star formation rates , which then constitute a minimum estimate .
however , upward corrections to this value are unlikely to be larger than a factor of two , since the observed rest - frame equivalent widths lie in the @xmath87 range close to the maximum values that are obtainable from ionization by a massive star population ( charlot & fall 1993 ) .
then , assuming case b recombination , we have @xmath88ly@xmath16 ) = 8.7@xmath88h@xmath16 ) ( brocklehurst 1971 ) , which using kennicutt s ( 1983 ) translation of @xmath76 from h@xmath16 luminosity , gives @xmath89ly@xmath16)/@xmath90 ergs s@xmath9 ) @xmath40 yr@xmath9 . in order to cross - calibrate to [ o ] fluxes at lower redshift we assume @xmath71(h@xmath16+[n ] ) = @xmath91([o ] ) based on the mean values of the ratio in both the gallego et al .
( 1995 ) and hawaii deep survey ( cowie et al . 1996 ) samples .
we also assume @xmath71(h@xmath16+[n ] ) = @xmath92(h@xmath16 ) ( kennicutt 1983 ) . in figure [ fig:5 ]
we compare the range of ly@xmath16 luminosities to the range of line luminosities in lower redshift objects .
the plot shows the quantity @xmath93 vs redshift , where h@xmath16+[n ] , [ o ] , and ly@xmath16 fluxes have been converted to h@xmath16 fluxes using the relationships above , and we have restricted ourselves to the imaging data in which the fluxes of the ly@xmath16 are well determined .
the comparison objects are drawn from the surveys of gallego et al .
1995 ( _ filled boxes _ ) , songaila et al .
1994 ( _ pluses _ ) , and the hawaii @xmath70 sample [ _ open boxes _ ( h@xmath16+[n ] ) , _ triangles _ ( [ o ] ) , and _ circles _ ( ly@xmath16 ) ] . the solid ( @xmath94 ) and dashed ( @xmath95 ) lines on fig .
[ fig:5 ] show the fluxes corresponding to stellar mass production rates of 10 @xmath40 yr@xmath9 , 1 @xmath40 yr@xmath9 , and 0.1 @xmath40 yr@xmath9 for @xmath96 .
maximum formation rates locally are around a few @xmath40 yr@xmath9 , rising to values of just over 10 @xmath40 yr@xmath9 above @xmath97 .
the ly@xmath16 fluxes at the higher redshifts are then consistent with this value or just slightly smaller depending on the extinction correction .
the cadis results ( thommes et al .
1998 ) would lie a factor of several times higher in flux than the two hawaii filter samples , but both keck spectroscopy and repeat fabry - prot observations have disproved the original @xmath98 candidate selection ( meisenheimer 1998 ) .
the minimum integrated star formation rates at @xmath79 and @xmath99 are 0.006 @xmath40 mpc@xmath18 yr@xmath9 and 0.01 @xmath40 mpc@xmath18 yr@xmath9 respectively for @xmath94 where the first value is slightly smaller than that given in paper i since agn - like objects are excluded .
both values are lower limits calculated in the absence of extinction and the @xmath99 value is based on a single field . within the substantial uncertainties of the as yet small samples ,
the results suggest that the star formation rates in the strong emission line population are constant or may possibly be increasing with redshift from @xmath14 , in contrast to color - based samples where the rate is declining at higher redshifts ( madau et al . 1996 , 1998 ) .
this is consistent with the broad expectation that as we move to higher redshifts and earlier stages of galaxy formation , an increasingly larger fraction of the star formation should be in strong ly@xmath16 emitters that correspond to the youngest galaxies .
we thank t. bida , r. goodrich , j. aycock , r. quick , and w. wack for their assistance in obtaining the observations , which were made possible by the lris spectrograph of judy cohen and bev oke .
we would also like to acknowledge helpful communications from k. meisenheimer .
this work was supported by the state of hawaii and by nasa grants go-5975.01 - 94a , go-5922.01 - 94a , go-6626.01 - 95a , and ar-6337.06 - 94a from space telescope science institute , which is operated by aura , inc .
, under nasa contract nas 5 - 26555 .
e.m.h . acknowledges a university research council seed money grant .
thanks the royal society for support . | we describe narrowband and spectroscopic searches for emission - line star forming galaxies in the redshift range 3 6 with the 10 m keckii telescope .
these searches yield a substantial population of objects with only a single strong ( equivalent width @xmath0 ) emission line , lying in the @xmath1 range .
spectra of the objects found in narrowband selected samples at @xmath2 and @xmath3 show that these very high equivalent width emission lines are generally redshifted ly@xmath4 at @xmath5 and 4.5 .
the density of these emitters above the 5@xmath6 detection limit of @xmath7 ergs @xmath8 s@xmath9 is roughly 15,000/@xmath10/unit @xmath11 at both @xmath5 and 4.5 .
a complementary deeper ( @xmath12 ergs @xmath8 s@xmath9 ) slit spectroscopic search covering a wide redshift range but a more limited spatial area ( @xmath13 ) shows such objects can be found over the redshift range @xmath14 , with the currently highest redshift detected being at @xmath15 .
the ly@xmath16 flux distribution can be used to estimate a minimum star formation rate in the absence of reddening of roughly @xmath17 mpc@xmath18 yr@xmath9 ( @xmath19 , @xmath20 ) .
corrections for reddening are likely to be no larger than a factor of two , since observed equivalent widths are close to the maximum values obtainable from ionization by a massive star population . within the still significant uncertainties ,
the star formation rate from the ly@xmath16selected sample is comparable to that of the color - break selected samples at @xmath21 , but may represent an increasing fraction of the total rates at higher redshifts .
this higher-@xmath11 population can be readily studied with large ground - based telescopes . |
it is generally recognized that the sagittarius dwarf spheroidal galaxy @xcite is currently disrupting under the strain of the milky way tidal field .
the stars ( and clusters , see * ? ? ?
* ; * ? ? ?
* ) lost by the sgr dsph remain coherently aligned along the orbital path of the galaxy , forming a huge filamentary structure , the sgr stream , that has now been observed over the entire sky ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and references therein ) .
therefore , the identification and definition of all the stellar components belonging to the sgr dsph is crucial not only to understand the star formation history ( sfh ) and chemical evolution of this keystone galaxy , but also to trace its remnants in the galactic halo .
the color magnitude diagram ( cmd ) of the sgr galaxy is dominated by the classical features of a metal - rich population , with a red and extended red giant branch ( rgb ) and a red clump ( rc ) of he - burning stars @xcite .
it is now generally recognized that this dominant population has a mean age of 4 - 6 gyr ( @xcite hereafter ls00 , @xcite , hereafter pap - i ) and a high mean metallicity ( * ? ? ?
* ; * ? ? ?
* ls00 , pap - i ) .
the apparent absence of old and metal poor stars ( oms ) in sgr was noted since the early studies and persists , in part , at the present day .
however , oms should be present , since ( a ) previous generations of stars are required to enrich the medium from which the dominant metal - rich population has formed and ( b ) the sgr dsph hosts at least two classical old and metal poor globular cluster ( ter 8 and m 54 , see * ? ?
* , respectively ) .
hence , we have observational evidence that oms were formed in sgr , even if just in globular clusters . unfortunately , the search for rgb stars bluer than the main rgb of sgr ( hence , presumably , more metal poor ) has been hampered by the strong contamination by foreground stars affecting the cmd .
therefore , the available spectroscopic surveys included only a negligible fraction of stars with @xmath2<-1 $ ] @xcite . a blue horizontal branch ( bhb )
sequence is easily recognizeable in the wide - field photographic cmd of @xcite and it can be considered a clear indication of the presence of oms in sgr : however this population has never been further observed in any successive sgr ccd survey .
some additional indirect clues of the existence of a metal - poor population in sgr exist in the literature , in particular : 1 .
@xcite surveyed an isolated field in sgr and found a star count excess in a region of the cmd similar in magnitude and color to the hb of ter 8 .
however , due to the few ( 23 ) sgr bhb candidates , this was considered just as a supporting evidence that old and metal poor stars exist in the field of sgr .
2 . @xcite analized the light curve of @xmath01800 rr lyrae in the sgr galaxy . from the location of the rrd variables in the petersen diagram he concluded that rr lyrae in sgr have metallicities in the range @xmath3@xmath4 ( see also ls00 ; * ? ? ?
since rr lyrae stars are thought to trace old populations ( 10 gyr ) , this result is a convincing proof that some old and metal poor stars are present in the sgr field population . according to @xcite ,
the central density of rr lyrae in sgr is @xmath5 stars / deg@xmath6 .
since we found ( pap - i ) @xmath02000 stars / deg@xmath6 belonging to the red clump , the rr lyrae should account for just @xmath06% of the whole hb population .
in this letter we use the cmd of a large area ( @xmath7 deg@xmath6 ) in the main body of the sgr dsph ( pap - i ) to show that a significant bhb population is indeed present in sgr .
we also show that the bhb is most likely the counterpart of the analogous feature found in the sgr stream by @xcite .
the database presented in pap - i contains v and i photometry of @xmath0490,000 sources down to @xmath8 in a @xmath7 deg@xmath6 field centered on the globular cluster m 54 , which coincides with the position of the maximum density peak of the sgr galaxy ( see pap - i for further details ) .
figure 1 shows a zoom of the hb region of the cmd , at different distances from the center of m 54 .
panel ( a ) refers to stars contained within the tidal radius of m 54 must always be intended as the tidal radius of the globular cluster m 54 . ]
( @xmath9 , where @xmath10 , * ? ? ?
as can be seen , although the galactic foreground stars ( vertical structure at v - [email protected] ) and the main sgr population ( red clump at @xmath12 ) are easily recognizable , the population of m 54 obviously dominates the cmd .
in particular the cluster rgb , ranging from ( v , v - i)@xmath0(21 , 0.9 ) up to ( 16.4 , 1.3 ) , and the well known bhb , at v - i0.4 from @xmath13 down to @xmath14 , are clearly evident ( see also ls00 and references therein ) . the bhb sequence of m 54 has been enclosed in a box that is repeated in all panels .
panel ( b ) shows the cmd of stars lying at r@xmath15r@xmath16 : as expected , while the rgb of m 54 disappears , the galactic foreground and the sgr population dominate the cmd at this distance from m 54 .
however , the most surprising feature is the unexpected presence of a conspicuous population of bhb ( see also panels ( c ) and ( d ) ) .
the bhb population ( and in particular its brightest bulk at [email protected] ) remains quite numerous even in the most external part ( @xmath18 ) of the surveyed region ( see panel ( d ) ) .
figure 1 thus demonstrates that a significant bhb population is indeed present in the main body of the sgr galaxy , as previously suggested by @xcite and @xcite .
the morphology of the bhb found in sgr is quite similar to that of m 54 , although the faintest part of the sequence ( [email protected] ) appears less populated with respect to m 54 . in order to further test the connection between the bhb population and the sgr galaxy , we have selected two different stellar samples ( shown in the left panel of figure 2 ) and we have compared their radial distributions .
the first sample contains stars belonging to the rc and therefore traces the distribution of stars belonging exclusively to the sgr galaxy .
the second contains stars belonging to the brightest part of the bhb only ( [email protected] ) to avoid spurious effects due to incompleteness and contamination by the blue plume .
the bhb sample should therefore trace the bhb of m 54 plus the contribution of the sgr bhb .
figure 3 shows the cumulative radial distribution of the two adopted test samples outside @xmath19 ( left panel ) and @xmath20 ( right panel ) . as can ben seen from both panels of figure 3 the two samples appear remarkably similar .
a kolmogorov - smirnov test provides a large probability ( @xmath1566% ) that the two distributions are extracted from the same parent population , supporting the idea that bhb stars outside m 54 actually belong to the sgr galaxy .
figure 4 shows instead the cumulative distribution of the two samples for @xmath21 , well inside the tidal radius of m 54 , but still far enough from the very core of the cluster , where crowding becomes severe .
the cumulative radial profile of m 54 is represented by a king model @xcite with @xmath22 and c=1.84 @xcite . from the inspection of the figure it is then evident that , even inside @xmath19 , the cumulative distribution of the bhb sample is significantly different from the star distribution in m 54 .
in fact , to reproduce the observed bhb sample distribution , we need to use a combined population .
figure 4 shows that the global distribution of bhb stars is nicely reproduced by a composite population with ( approximately ) @xmath030% of stars contributed by m 54 and @xmath070% by sgr .
however , a good agreement can also be achieved by varying the relative fractions by @xmath010% .
in the previous section we report convincing evidence that a substantial population of bhb is present in the main body of sgr .
the existence of this population was missed by all previous photometric ccd surveys , probably due to an unsufficient field of view .
this recovery can shed light on different aspects of the structure and the history of this disrupting stellar system .
the presence of a well - populated bhb in sgr implies the presence of a substantial population of oms in this galaxy .
in fact , bhb have been , up to now , observed only in old , metal - poor stellar populations .
the only exceptions are two metal - rich and very dense globular clusters belonging to the bulge @xcite .
however , bhb stars appear spatially segregated toward the center of these clusters , suggesting a dynamical origin for these stars @xcite . due to the extremely low stellar density conditions
, this kind of formation channel appears quite unlikely in the case of the sgr galaxy .
thus , with a morphology that is very similar to that of the old , metal - poor clusters associated to the sgr : ter 8 ( [ fe / h]@xmath23 ) and m 54 ( [ fe / h]@xmath24 ) , the bhb sequence of sgr is more likely composed of genuine oms stars .
moreover , although it is often difficult to place precise constraints on age and metallicity from the study of the hb morphology alone , we can anyway put some limits to the age and metallicity of the bhb population by observing that stars bluer than the rr lyrae should be at least as old and metal - poor as the rr lyrae themselves . in particular rr lyrae stars are known to trace old populations ( @xmath2510 gyr ) .
hence , by combining this with the results by @xcite , it can be concluded that _ the sgr bhb population is composed by old and metal - poor stars , with ages greater than @xmath010 gyr and metallicities lower than [ fe / h]@xmath11 - 1.3_. we select all the rc stars with @xmath26 ( lower selection box in the right panel of figure 2 ) .
then we estimate the level of foreground contamination by counting stars within a box adjacent to the rc sample box ( upper selection box in the right panel of figure 2 ) having the same size , and we obtain n@xmath27@xmath01000 stars/0.95 deg@xmath6 ( we consider all stars with @xmath26 , i.e. a total area of 0.95 deg@xmath6 ) .
we use this estimate to clean the rc sample , obtaining a final rc density of n@xmath28@xmath02000 stars/0.95 deg@xmath6 . in the same region , we counted stars belonging to the bright bhb sample defined in figure 2 , obtaining n@xmath29147 stars/0.95 deg@xmath6 . finally , from the work of @xcite , the expected number of rr lyrae is n@xmath30@xmath11130 stars/0.95 deg@xmath6 .
we conclude that _ the fraction of old and metal - poor stars in the sgr galaxy is @xmath012% . _ therefore , while the most important episodes of star formation in the sgr galaxy occurred between 4 and 6 gyr ago ( pap - i , corresponding , @xmath31 and @xmath32 has been assumed . ] to @xmath33 ) from a quite metal - rich ( [ m / h]-0.6 ) interstellar medium , a non negligible fraction of metal poor stars were already in place more than 10 gyr ago ( @xmath342 ) .
@xcite estimated a larger , but still compatible , fraction of oms : @xmath3530% .
however it has to be noted that our estimate should be considered a lower limit since a larger fraction can be obtained if fainter bhb are taken into account .
recently @xcite , using the early data release of the sloan digital sky survey , sdss , report on the discovery of an unexplained stellar population in a region of the sky named s341 + 57 - 22.5 , in the direction @xmath36 .
the authors suggest that the unknown population is part of the sgr stream since ( 1 ) s341 + 57 - 22.5 lies on the predicted sgr s orbital path ( according to * ? ? ?
* ; * ? ? ?
* ) and ( 2 ) the cmd is quite similar to that of the sgr galaxy published by @xcite , with the notable exception of a bhb sequence which was lacking in any previous cmd of sgr ( with the only exception of the photographic cmd by * ? ? ?
we used the same equations that @xcite applied to the ( v , v - i ) photometry of @xcite to transform our photometry into the sdss photometric system , adopting the same distance and reddening as well .
the bhb presented in this letter is then located in the cmd at @xmath37 , a position which is fully compatible with the bhb sequence discovered by @xcite .
thus , the bhb population found by @xcite in the sgr stream corresponds to the bhb population found in the sgr galaxy .
this implies that the disruption of the sgr dsph provided a sizeable contribution to the assembly of the galactic halo not only in the form of metal - rich stars ( traced by the m giants studied by * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) but also providing a substantial amount of old and metal - poor stars that constitute the typical halo population . 300 bellazzini , m. , ferraro , f.r . ,
buonanno , r. , 1999a , , 304 , 633 bellazzini , m. , ferraro , f.r . ,
buonanno , r. , 1999b , , 307 , 619 bellazzini , m. , ferraro , f.r .
& ibata , r. , 2003a , aj , 125 , 188 bellazzini , m. , ibata , r. , ferraro , f.r . , & testa , v. , 2003b , , in press ( astro - ph/0305107 ) bonifacio , p. , & caffau , e. , 2003 , , 339 , 1183 cacciari , c. , bellazzini , m. , & colucci , s. , 2002 , in extragalactic star clusters , iau symposium 207 , d. geisler , e.k .
grebel , and d. minniti eds . , san francisco : astronomical society of the pacific , p. 168 cseresnjes , p. , 2001
, a&a , 375 , 909 king , i.r . , 1962 , , 67 , 471 kundu , a. , et al . , 2002 ,
apj , l125 ibata , r.a . , irwin , m.j . , & gilmore , g. , 1994 , nature , 370 , 194 ibata , r. a. , gilmore , g. , & irwin , m. j. 1995 , , 277 , 781 ibata , r.a . ,
wyse , r.f.g . ,
gilmore , g. , irwin , m.j . , & suntzeff , n.b .
, 1997 , , 113 , 634 ibata , r.a . , &
lewis , g.f .
, 1998 , , 500 , 575 ( il98 ) ibata , r.a . ,
lewis , g.f . , irwin , m. , totten , e. , & quinn , t. , 2001a , , 551 , 294 ibata , r.a . , irwin , m. , lewis , g.f . , & ibata , r.a . , lewis , g.f .
, irwin , m. & cambrsy , l. , 2002 , mnras , 332 , 921 ivezic , z. , et al .
, 2000 , , 120 , 963 layden , a.c . ,
ritter , l.a .
, welch , d.a . , & webb , t.m.a . , 1999 , , 117 , 1313 layden , a.c . ,
& sarajedini , a. , 2000 , , 119 , 1760 ( ls00 ) majewski , s.r . ,
skrutsie , m.f . ,
weinberg , m.d .
, ostheimer , j.c . , 2003 , , submitted ( astro - ph/0304198 ) marconi , g. , buonanno , r. , castellani , m. , iannicola , g. , molaro , p. , pasquini , l. & pulone , l. , 1998 , , 330 , 453 martnez - delgado , d. , aparicio , a. , gmez - flechoso , m.a .
, & carrera , r. , 2001 , , 549 , l199 mateo , m. , udalski , a. , szymansky , m. , kaluzny , j. , kubiak , m. , & krzeminski , w. , 1995 , , 110 , 1141 monaco l. , ferraro , f.r . ,
bellazzini , m. , pancino , e. , 2002 , , 578 , l47 ( pap - i ) montegriffo , p. , bellazzini , m. , ferraro , f.r . , martins , d. , sarajedini , a. , & fusi pecci , f. , 1998 , , 294 , 315 newberg , h.j . , et al . , 2002 , , 569 , 245 ( n02 ) rich , r.m .
, et al . , 1997 , , 484 , l25 sarajedini , a. , layden , a.c . , 1995 , , 109 , 1086 smecker - hane , t.a
. , & mcwilliam , a. , 2002 , , submitted ( astro - ph/0205411 ) trager , s. c. , king , i. r. , & djorgovski , s. 1995 , , 109 , 218 yanny , b. , et al . , 2000 , , 540 , 825 | we report on the recovery of a blue horizontal branch ( bhb ) population belonging to the sagittarius dwarf spheroidal galaxy ( sgr ) .
the sequence is clearly identified in the ( v , v - i ) color magnitude diagram ( cmd ) obtained for about 500,000 stars in the region of the globular cluster m 54 .
the bhb morphology is similar to the analogous sequence in m 54 , but it is unambiguously associated with sgr since _ ( i ) _ it is detected well outside the main body of the cluster , up to more than 5 tidal radii from the cluster center and _ ( ii ) _ the bhb stars follow the radial distribution of the other stellar populations of sgr .
this finding finally demonstrates that the sgr galaxy hosts a significant ( of the order of @xmath010% ) old and metal - poor stellar population ( [ fe / h]-1.3 ; age 10 gyr ) , similar to that of its oldest clusters ( m 54 , ter 8) .
we also show that the sgr bhb sequence found here is the counterpart of the analogous feature observed by newberg et al .
( 2002 ) in the sgr stream , in a field more than 80@xmath1 away from the center of the galaxy . |
binary neutron stars are very interesting objects . from the observational point of view
, we will have a chance to get new eyes for the universe by detecting gravitational waves in the first or second decade of this century .
it is highly possible that the first signal may be that from compact binary stars , such as binary neutron stars , a black hole neutron star binary system , or binary black holes . on the other hand , theoretically , our understanding of evolution of compact binary stars is far from complete because it is considerably difficult to treat the `` 2-body '' problem from a state with a wide separation to a merging stage consistently in the framework of general relativity .
however , recent investigations have found a new approach to this problem . since
the time scale of the orbital change due to gravitational wave emission is rather long compared with the orbital period except for in the final few milliseconds of the coalescing stage , we can neglect gravitational wave emission for most stages of evolution . in other words , we can treat the system in `` quasiequilibrium '' ( see e.g. , @xcite ) .
following this idea , several groups have obtained quasiequilibrium sequences of binary neutron stars @xcite .
most of them adopted the assumption that the spatial part of the metric is conformally flat ( the conformally flat condition : hereafter , cfc ) @xcite .
for axisymmetric rotating polytropes results of the scheme with the cfc were compared with those obtained by the numerically exact code and found to be reasonably accurate @xcite . however , since there are no exact numerical solutions for binary configurations , one could not know the accuracy of the results obtained by the scheme with the cfc .
therefore , it is desirable to develop different schemes from that with the cfc and to compare results of different schemes for nonaxisymmetric configurations .
as one of those alternatives , in the previous paper , we developed a new numerical scheme to obtain quasiequilibrium structures of nonaxisymmetric compact stars as well as the space time around those stars in general relativity and obtained quasiequilibrium sequences of synchronously rotating binary polytropes @xcite . in that scheme ,
the einstein equations are solved directly without assuming the cfc .
the obtained results , however , could not be compared with those of @xcite because different polytropic relations were used . in this paper , we have used the same polytropic equation as that used in @xcite and computed quasiequilibrium sequences of synchronously rotating polytropes with a constant rest mass .
therefore , we can directly compare our results with those with the cfc .
in actual computations , we have improved our numerical scheme and succeeded in making our scheme more robust .
as mentioned in introduction , we can neglect the effect of gravitational wave emission for almost all stages except for the last few milliseconds of the coalescence of binary systems and choose a rotating frame in which the system is in a stationary state . in this paper
, we use the units of @xmath4 except for in the final section and we will make the following assumptions ( see , usui , ury , and eriguchi @xcite for a more detail ) : ( 1 ) : : we deal with a binary system which consists of two stars of equal masses in a circular orbit . ( 2 ) : : the binary star system is in a stationary state in the rotating frame with the orbital angular velocity @xmath5 .
( 3 ) : : axes of spins of two stars and that of the orbital motion are parallel to one another .
a schematic figure of the system is shown in fig . [ coordinate system ] .
( 4 ) : : spins of two stars are synchronized to the orbital motion .
each star is rigidly rotating with the angular velocity @xmath5 if seen from a distant place .
( 5 ) : : the matter of the star is a perfect fluid and the equation of state for the matter is assumed to be polytropic : @xmath6 where @xmath7 , @xmath8 , @xmath9 , @xmath10 , and @xmath11 are the pressure , the rest mass density , the energy density , the polytropic index , and the polytropic constant , respectively .
in addition to the assumptions mentioned above , we further assume the following form for the metric in the spherical coordinates @xmath12 : @xmath13 where @xmath14 , @xmath15 , @xmath16 , and @xmath17 are the metric potentials which are functions of @xmath18 , @xmath19 , and @xmath20 .
although this form of the metric becomes exact for stationary axisymmetric configurations , it is not exact for nonaxisymmetric configurations .
therefore , it must be extended to a more general form in the future .
the einstein equations with appropriate boundary conditions can be written down for the metric form eq .
( [ metric ] ) , and are transformed to integral representations by using the green function for the laplacian in the flat space ( see @xcite ) .
here we solve four equations for the metric potentials , @xmath14 , @xmath15 , @xmath16 , and @xmath17 , and the rest of the einstein equations are not used . in the concrete , we do not consider @xmath21 , @xmath22 , @xmath23 , and @xmath24 components of the einstein equations ( see @xcite ) .
we can check how these equations are satisfied or violated after obtaining the solutions to the rest of the einstein equations . in actual computations , even for highly relativistic configurations , these equations are satisfied quite well , to within @xmath25 , if we estimate the relative errors by the following quantity : @xmath26 where @xmath27 and @xmath28 are the ricci tensor , the energy momentum tensor and the trace of the energy momentum tensor , respectively .
it must be noted that for nonaxisymmetric quasiequilibrium configurations it is _ impossible _ to satisfy all components of the einstein equations because there can not exist `` exact '' quasiequilibrium configurations as far as the asymptotic flat condition is assumed .
the method of calculation is almost the same as that used in the previous paper @xcite except for several improvements .
we will briefly summarize the scheme and explain the improved changes in some detail .
we solve the hydrostatic equation and the metric potentials iteratively by adopting the hscf ( hachisu s self - consistent field ) method @xcite .
in particular , as mentioned before , the einstein equations for the metric are transformed into the integral equations by constructing the laplacian in the flat space . in this procedure
we have to add the following derivative to both sides of the equation : @xmath29 this term may cause the iteration diverge because , once a certain amount of numerical errors is introduced by chance , this error may not decrease but increase iteration by iteration near the coordinate center . to avoid this problem ,
we have shifted the coordinate center downwards along the @xmath30-axis as fig .
[ shifted coordinate system ] .
numerical computations are carried out only in the upper half of the whole space , i.e. , above the equatorial plane of the binary star systems , and quantities in the lower half are computed by making use of the symmetry of the physical quantities about the equatorial plane .
consequently , it is noted that 25% of the whole grid points is used in the actual calculations . as shown later , this scheme works nicely .
we have used @xmath31 grid points for ( @xmath32 , @xmath33 , @xmath34 ) , where @xmath35 , and @xmath36 means the distance from the orbital axis to point b ( see fig . [ coordinate system ] ) .
since there exist no exact equilibrium or stationary states for real compact binary stars with a flat space at infinity , it is difficult to define exactly the conserved quantities for nonaxisymmetric configurations in general relativity . however , for an asymptotically flat spacetime , we can define the _ approximate _ total angular momentum , @xmath37 , and the _ approximate _ total gravitational mass , @xmath38 , as : @xmath39r^2\sin\theta dr d\theta d\varphi \ , \\ m_{\rm tot } & = & \int e^{2\alpha+\beta+\nu}\left\{(\varepsilon + p)\frac{1+v^2}{1-v^2}+2p\right\ } r^2\sin\theta dr d\theta d\varphi\nonumber\\ & & { } + \int 2r\sin\theta e^{2\alpha+2\beta}(\varepsilon + p)\frac{v\omega}{1-v^2}r^2\sin\theta dr d\theta d\varphi \ , \end{aligned}\ ] ] where @xmath40 is the velocity of the matter .
these expressions can be obtained by making use of the behaviors of the metric functions at flat infinity as follows ( see @xcite for a more detail ) : @xmath41 notice that these expressions are not general forms and different from the adm quantities used in baumgarte _
et al . _
@xcite . to compare our results with those with the cfc
, it would be desirable to employ the same definitions of physical quantities as theirs . for conformally flat situations
, the adm mass can be easily evaluated by using the matter contribution in addition to the contribution of the extrinsic curvatures as seen in baumgarte _
et al . _
@xcite . with our choice of the metric ( eq . ( [ metric ] ) ) , however , it is difficult to use the same definition because we do not use the conformal factor
. therefore we follow the scheme used by bardeen @xcite and derive the total mass and the total angular momentum by considering the asymptotic behavior of the 4-metric . in the previous paper @xcite ,
the expression of the angular momentum ( eq.(48 ) in @xcite ) was incorrect .
numerical values of the angular momentum in @xcite contain about @xmath42 errors at maximum due to that wrong expression . on the other hand , for the rest mass
, it is natural to define as : @xmath43 since we use the same polytropic relation as that used by baumgarte _
et al . _
@xcite , we can choose the same nondimensional quantities as they did by using the polytropic constant @xmath11 and the polytropic index @xmath10 as : @xmath44 where @xmath45 , @xmath46 and @xmath47 are the gravitational mass , the rest mass and the angular momentum of a component star and quantities with `` bar '' mean nondimensional ones . for polytropes , the equation of state
is fixed by choosing one parameter , constant @xmath11 , in addition to the polytropic index @xmath10 .
after the equation of state is fixed , we need to specify two more parameters to determine one rotating equilibrium configuration : one represents the strength of gravity and the other for the rotation . in our formulation , we choose the following two parameters : ( 1 ) : : the maximum energy density , @xmath48 .
( 2 ) : : the ratio of the shortest distance @xmath49 ( distance from the orbital axis to point a , see fig . [ coordinate system ] ) to the largest distance @xmath36 ( distance from the orbital axis to point b , see fig .
[ coordinate system ] ) , @xmath50 : @xmath51 + this quantity @xmath50 indirectly specifies the rotational state and is related to the separation distance between the two stars as @xmath52 , where @xmath53 is the distance between the geometrical centers of the stars .
since the quantity @xmath11 does not appear in the basic equations due to our normalization , i.e. , eqs .
( [ nondimensional m])([nondimensional j ] ) , we can treat the problem by specifying only three parameters , @xmath48 ( or nondimensional quantity @xmath54 ) and @xmath50 , in addition to the polytropic index @xmath10 .
in order to consider realistic evolutions of compact binary star systems , it would be important to find conserved quantities which characterize the evolution .
although , during evolutions of binary neutron star systems , the gravitational mass and the angular momentum are lost from the system , the baryon number of the system should be conserved because we do not consider the mass loss or the roche lobe overflow .
therefore a sequence of a constant rest mass ( baryon mass ) can be considered to provide an evolutionary track .
even if the rest mass of the star is specified , one can not follow an evolutionary sequence of binary stars .
one needs to know the following two things : 1 ) the change of the equation of state and 2 ) the change of the velocity field of the stars .
concerning the first point , as far as one uses the polytropic relations , it is impossible to know the change of the equation of state .
thus we adopt the same approximation as baumgarte _
_ did , i.e. , the polytropic index @xmath10 and the constant @xmath11 are fixed during binary evolutions @xcite . concerning the value of @xmath11
, it may be a reasonable assumption because @xmath11 means the entropy in the newtonian limit and the evolution can be treated as an almost adiabatic process , i.e. , the timescale of viscous heating is much longer than evolutionary time scale due to the back reaction of gravitational wave emission . as for the second point
, it is widely believed that binary neutron star systems evolve irrotationally @xcite . however , since the purpose of the present paper is to present an alternative method of solving binary neutron stars without employing the cfc , we assume that the system evolves by keeping synchronous rotation , although its evolution is not realistic .
although our final goal in this paper is to obtain evolutionary sequences of constant rest masses for the specified polytropic relation , it is not easy to follow constant rest mass sequences because the rest mass is not a local quantity but a global quantity which can be computed after the structure of the binary star is obtained .
thus , we construct evolutionary sequences of constant rest masses by the following procedure : ( 1 ) : : fix the polytropic index @xmath10 .
( 2 ) : : calculate sequences by changing @xmath54 for a fixed value of @xmath50 . after obtaining the structure ,
compute global physical quantities such as the rest mass , the gravitational mass , and the angular momentum .
( 3 ) : : repeat step ( 2 ) by changing the value of @xmath50 . ( 4 ) : : specify one value of the rest mass and interpolate physical quantities with different values of @xmath50 for the specified rest mass by using quantities obtained in steps ( 2)-(3 ) .
thus constant rest mass sequences are obtained .
( 5 ) : : change the rest mass and repeat step ( 4 ) and construct sequences for different strength of gravity .
hereafter , we choose the compactness , @xmath55 , as the parameter of the strength of gravity . here
@xmath56 and @xmath57 are the gravitational mass and the schwarzschild radius of a spherical star with the same value of the rest mass as that of the sequence which we consider . in fig .
[ dens - mass n=1.0 ] , the nondimensional rest mass of each star is plotted against the nondimensional central energy density .
each curve represents a sequence along which the separation @xmath50 is fixed but the energy density @xmath54 is varied .
seven curves correspond to the sequences of @xmath58 ( contact ) , @xmath59 , @xmath60 , @xmath61 , @xmath62 , @xmath63 , and @xmath64 . from this figure
, we can see that as the central energy density increases , the rest mass increases until a certain critical point .
however , as the density increases further , the mass begins to decrease because of the general relativistic effect .
the curves at the upper side correspond to the binary sequences with smaller separations and the uppermost curve denotes the sequence at the contact phase . for larger separations , each star can be considered to approach the spherical configuration . by using the information in this figure
, we can consider evolutionary sequences with constant rest masses because horizontal lines can be regarded as such sequences .
thus , we choose four evolutionary sequences of four constant rest masses which correspond to four different compactnesses , i.e. , @xmath65 , @xmath60 , @xmath61 , and @xmath62 .
the energy density contours for selected models at the contact phase are shown in fig .
[ contours ] . in this figure ,
equidensity contours are drawn for several @xmath10 and @xmath55 . in fig .
[ j - omega n=1.0 ] , the nondimensional angular velocity is plotted against the nondimensional angular momentum .
four curves correspond to sequences with the compactness mentioned above .
as the angular momentum decreases due to gravitational wave emission , the binary system evolves to the upper - left direction along the curve .
if the curve has a turning point , the system becomes secularly unstable at that point .
as seen from this figure , for @xmath1 polytropes , it is very difficult to tell whether turning points appear or not , because turning points , if exist , seem to coincide with the contact states . in this same figure , the results of baumgarte _ et al . _
@xcite are also plotted by using several different types of symbols .
if we assume that the turning points coincide with the contact points , we can see almost the same tendency about the occurrence of secular instability , although the values are not the same . while baumgarte _
et al . _
@xcite used the adm quantities , we have introduced our approximate definitions as mentioned in sec .
[ physical quantities ] .
although these quantities coincide in the newtonian limit , the slight differences appear for highly relativistic , highly nonlinear configurations .
[ q - dens n=1.0 ] shows the nondimensional central energy density @xmath54 separation parameter @xmath50 relation . from this figure
, we can see that the central energy density decreases as the orbit shrinks , i.e. , as the evolution proceeds .
the central energy density seems to decrease as the separation become larger , too .
however , this may not be real because , as mentioned before , a small number of grid points cover the matter region and values can not be highly accurate .
thus , there is no tendency to collapse individually prior to merging , which was suggested by wilson _
_ @xcite . in figs .
[ dens - mass n=1.5 ] and [ dens - mass n=0.5 ] , the nondimensional rest mass of each star is plotted against the nondimensional central energy density for @xmath2 and @xmath0 sequences , respectively .
each curve represents a sequence along which the separation @xmath50 is fixed but the energy density @xmath54 is varied .
the tendency is the same as that for @xmath1 sequences . as for the constant rest mass sequences , we show three selected sequences with different compactness , i.e. , @xmath66 , @xmath67 , and @xmath68 for @xmath2 polytropes , and four sequences , i.e. , @xmath65 , @xmath60 , @xmath62 and @xmath64 for @xmath0 polytropes .
they are shown in figs .
[ j - omega n=1.5 ] and [ j - omega n=0.5 ] . as in fig .
[ j - omega n=1.0 ] , the nondimensional angular velocity is plotted against the nondimensional angular momentum . as seen from these figures , for @xmath69 polytropes
, the sequences terminate at the contact phases without encountering the turning points .
it implies that the binary system with a soft equation of state evolves to contact states stably . on the other hand , for stiff polytropes , @xmath3 , the turning points appear before the contact phases so that the binary system becomes unstable .
in this paper we have solved quasiequilibrium sequences of synchronously rotating binary star systems in general relativity without assuming the cfc .
we have constructed the constant rest mass sequences and shown that for stiff equations of state ( @xmath3 ) , evolutionary curves have turning points so that synchronous rotation of the system breaks down at that point and that the internal motion will be excited .
our results can be compared with those of baumgarte _ et al . _ who employed the cfc @xcite .
quantitatively , there are some differences between two results as seen from figs .
[ j - omega n=1.0 ] and [ j - omega n=1.5 ] .
these differences may come from different choices of the metric .
however , it should be noted that qualitative features are very similar , i.e. , the dependency on the polytropic index of the appearance of the turning points and so on .
therefore , although it is hard to give exact values of the angular velocity and/or the angular momentum at the turning points from quasiequilibrium approach , the occurrence of the instability could be correctly predicted .
nevertheless , since there are no exact solutions for the binary neutron star systems , we should keep in mind that there is a possibility that both of the two results might not represent the exact solutions . in figs .
[ dj - dm n=0.5 ] and [ dj - dm n=1.0 ] , the nondimensional gravitational mass and the nondimensional angular momentum of each star are plotted against the nondimensional angular velocity . from these figures
, it can be seen that turning points , i.e. , the minima of each value , of two curves coincide .
it implies that secular stability of binary systems can be found by investigating either the gravitational mass or the angular momentum .
this is a nice feature that agrees the requirement between the change of the gravitational mass and that of the angular momentum as follows : @xmath70 where @xmath71 and @xmath72 are the changes of the gravitational mass and the angular momentum of two configurations with the same rest mass , respectively .
this relation can be reduced from the first law of thermodynamics , which is shown below , for the binary systems for which the rest mass , entropy , and vorticity of each fluid element are conserved @xcite : @xmath73 where @xmath74 means the half of the total energy of the system .
it should be noted that these requirements can be checked if we can obtain highly accurate models . as seen from figs .
[ dj - dm n=0.5 ] and [ dj - dm n=1.0 ] , changes of the gravitational mass and the angular momentum are three or four orders of magnitude smaller than the corresponding quantities .
unfortunately , since we can not insist that our values have such high accuracy , we do not show our results here . as seen from figs .
[ j - omega n=1.0 ] , [ j - omega n=1.5 ] and [ j - omega n=0.5 ] , the ranges of the values of @xmath75 and @xmath76 are considerably wide for the values of @xmath10 . even for the sequences with the same @xmath10 , the values of @xmath75 and @xmath76 range widely .
thus it is not easy to understand the effects of the strength of gravity and/or the equation of state . in order to see the features of the evolutionary sequences at a glance
, we introduce the following two nondimensional quantities , one of which can be considered to represent the angular velocity and the other of which corresponds to the angular momentum : @xmath77 where @xmath78 here , @xmath79 means the radius of the star on the major axis measured in the schwarzschild - like coordinate .
our coordinate system in this paper is a kind of isotropic one and so @xmath79 is defined as follows : @xmath80 the meaning of these quantities , @xmath81 and @xmath82 , can be roughly understood if we consider a system in newtonian gravity which consists of two identical rigid spheres of uniform density in a contact phase .
for such a system , @xmath83 and @xmath84 .
another property of these quantities can be seen from the definition of the normalization factors , eqs .
( [ normalization j ] ) and ( [ normalization omega ] ) . in these expressions , the differences originating from the different strength of gravity are renormalized " by introducing the term related to the quantity @xmath55 .
thus we will call @xmath85 a renormalized angular velocity and @xmath86 a renormalized angular momentum . in figs .
[ j - omega new ] , the renormalized angular velocity @xmath85 is plotted against the renormalized angular momentum @xmath86 for several sequences of @xmath10 and @xmath55 . as seen from this figure ,
the values of @xmath85 and @xmath86 for all evolutionary sequences with constant rest masses are scaled to values around unity . for the newtonian sequences , the position of contact phases for smaller values of @xmath10 approaches @xmath87 but never reaches that point because configurations are not rigid bodies and deformed from spheres by the tidal force from the companion star .
several characteristic features can be seen in this figure .
first , if we compare the sequences with the same value of @xmath55 , sequences with stiffer equations of state locate at the upper - right region .
this can be explained as follows .
if we choose models which have the same values of the gravitational mass and the angular velocity , the radii are the same so that the inertial moment is larger for the stiffer polytropes .
it implies that the angular momentum is larger for stiffer equations of state . concerning the value of the renormalized angular velocity at the contact stage ,
the gravitational force is stronger for the stiffer polytropes because of the distribution of the matter inside the star .
thus larger angular velocity is required for configurations with stiffer equations of state .
second , if we compare the sequences with the same value of @xmath10 , more relativistic sequences locate at the larger values of @xmath86 .
this is explained as follows .
if we choose the models which have the same values of @xmath10 , @xmath45 and @xmath85 , we obtain the following relation : @xmath88 where @xmath89 is the mass average of the quantity @xmath90 defined as follows : @xmath91 here @xmath92 is the mass element of the configuration .
since , in general , the change of the averaged quantity is smaller than the change of the quantity itself , the change of @xmath86 is affected mainly by the change of @xmath79 which decreases as @xmath55 increases . therefore the value of @xmath86 increases and the curves are shifted towards right in the plane .
it should be noted that if we consider sequences with the same value of @xmath55 but different values of @xmath10 , differences due to different values of @xmath10 are amplified for configurations with the larger value of @xmath55 . since real neutron stars can not be approximated by a single polytropic relation all through the whole star , the evolutionary sequences can not be approximated by the assumptions adopted in this paper , i.e. , the assumption that @xmath10 and @xmath11 are conserved .
therefore , in order to get information about real evolutions , we need to construct evolutionary sequences with realistic equations .
+ we would like to thank dr .
kji ury for his helpful discussions .
fu is a research fellow of the japan society for the promotion of science ( jsps ) and is grateful to jsps for the financial support .
this work was partially supported by the grant - in - aid for scientific research ( c ) of jsps ( 12640255 ) .
wilson , g.j .
mathews , and p. marronetti , phys .
d * 54 * , 1317 ( 1996 ) .
baumgarte , g.b .
cook , m.a .
scheel , s.l .
shapiro , and s.a .
teukolsky , phys .
d * 57 * , 7299 ( 1998 ) .
s. bonazzola , e. gourgoulhon , and j .- a .
marck , phys .
lett . , * 82 * , 892 ( 1999 ) .
k. ury , and y. eriguchi , phys .
d. * 61 * , 124023 ( 2000 ) .
f. usui , k. ury , and y. eriguchi , phys .
d. * 61 * , 24039 ( 2000 ) .
cook , s.l .
shapiro , and s.a .
teukolsky , phys .
d * 53 * , 5533 ( 1996 ) . i. hachisu , astrophys . j. suppl .
* 62 * , 461 ( 1986 ) .
h. komatsu , y. eriguchi , and i. hachisu , mon . not .
. soc . * 237 * , 355 ( 1989 ) .
bardeen , in black holes , eds . c. dewitt and b.s .
dewitt , 241 ( 1973 ) .
kochanek , astrophys .
j. * 398 * , 234 ( 1992 ) .
l. bildsten and c. cutler , astrophys .
j. * 400 * , 175 ( 1992 ) .
friedman , k. ury , and m. shibata , preprint ( gr - qc/0108070 ) .
axis is set along the line joining the two centers of mass of the two stars and the origin is set at their midpoint .
the @xmath30 axis is along the rotation axis .
points a , b are set at the intersections of the surface of the star and the @xmath93 axis .
the inner intersection is point a and the outer one is point b.[coordinate system],scaledwidth=50.0% ] is plotted against nondimensional central energy density @xmath54 for the quasiequilibrium @xmath1 polytropic binary star systems with the separation @xmath94 ( contact ) , @xmath59 , @xmath95 , @xmath61 , @xmath96 , @xmath63 , and @xmath97 .
the curves for binaries with smaller separations locate at upper side , and the uppermost curve denotes a sequence with @xmath94 .
thin solid curve denotes a sequence of spherical stars with the same polytropic index .
four horizontal lines labeled by @xmath65 , @xmath60 , @xmath61 , and @xmath62 indicate the constant rest mass sequences discussed in fig .
[ j - omega n=1.0 ] and fig .
[ q - dens n=1.0].[dens - mass n=1.0],scaledwidth=75.0% ] ccccc + @xmath98&&&&@xmath65 + -@xmath30 plane for models at the contact phase : ( a ) @xmath0 , ( b ) @xmath1 , and ( c ) @xmath2 .
the difference between two adjacent curves is 1/10 of the maximum energy density .
the units of the distance is @xmath36 . for each @xmath10 ,
contours for two values of the strength of gravity are shown : one for less relativistic models , and the other for highly relativistic models which correspond to the models with almost the maximum mass of the sequence .
[ contours ] , title="fig:",scaledwidth=30.0% ] & & & & -@xmath30 plane for models at the contact phase : ( a ) @xmath0 , ( b ) @xmath1 , and ( c ) @xmath2 .
the difference between two adjacent curves is 1/10 of the maximum energy density .
the units of the distance is @xmath36 . for each @xmath10 ,
contours for two values of the strength of gravity are shown : one for less relativistic models , and the other for highly relativistic models which correspond to the models with almost the maximum mass of the sequence .
[ contours ] , title="fig:",scaledwidth=30.0% ] + + + @xmath99&&&&@xmath65 + -@xmath30 plane for models at the contact phase : ( a ) @xmath0 , ( b ) @xmath1 , and ( c ) @xmath2 .
the difference between two adjacent curves is 1/10 of the maximum energy density .
the units of the distance is @xmath36 . for each @xmath10 ,
contours for two values of the strength of gravity are shown : one for less relativistic models , and the other for highly relativistic models which correspond to the models with almost the maximum mass of the sequence .
[ contours ] , title="fig:",scaledwidth=30.0% ] & & & & -@xmath30 plane for models at the contact phase : ( a ) @xmath0 , ( b ) @xmath1 , and ( c ) @xmath2 .
the difference between two adjacent curves is 1/10 of the maximum energy density .
the units of the distance is @xmath36 . for each @xmath10
, contours for two values of the strength of gravity are shown : one for less relativistic models , and the other for highly relativistic models which correspond to the models with almost the maximum mass of the sequence .
[ contours ] , title="fig:",scaledwidth=30.0% ] + + + @xmath100&&&&@xmath65 + -@xmath30 plane for models at the contact phase : ( a ) @xmath0 , ( b ) @xmath1 , and ( c ) @xmath2 .
the difference between two adjacent curves is 1/10 of the maximum energy density .
the units of the distance is @xmath36 . for each @xmath10 ,
contours for two values of the strength of gravity are shown : one for less relativistic models , and the other for highly relativistic models which correspond to the models with almost the maximum mass of the sequence .
[ contours ] , title="fig:",scaledwidth=30.0% ] & & & & -@xmath30 plane for models at the contact phase : ( a ) @xmath0 , ( b ) @xmath1 , and ( c ) @xmath2 .
the difference between two adjacent curves is 1/10 of the maximum energy density .
the units of the distance is @xmath36 . for each @xmath10 ,
contours for two values of the strength of gravity are shown : one for less relativistic models , and the other for highly relativistic models which correspond to the models with almost the maximum mass of the sequence .
[ contours ] , title="fig:",scaledwidth=30.0% ] is plotted against nondimensional angular momentum @xmath76 for the sequences of constant rest mass polytropic binary systems with @xmath1 .
four curves correspond to sequences with the compactnesses @xmath65 , @xmath60 , @xmath61 , and @xmath62 .
terminal point of each curve at higher values of the angular velocity corresponds to the configuration at the contact phase .
the results of baumgarte _ et al . _
@xcite are marked by different symbols and turning points in their results are denoted by the same but larger symbols .
[ j - omega n=1.0],scaledwidth=75.0% ] is plotted against the separation @xmath50 ( defined as eq .
( [ q define ] ) ) for the sequences of constant rest mass polytropic binary systems with @xmath1 .
four curves correspond to sequences with the compactnesses @xmath65 , @xmath60 , @xmath61 , and @xmath62 .
the quantity @xmath101 is the central energy density of a spherical model.[q - dens n=1.0],scaledwidth=75.0% ] but for @xmath2 .
three curves correspond to sequences with the compactness @xmath66 , @xmath67 , and @xmath68 .
different symbols denote the results of baumgarte _ et al .
_ @xcite .
[ j - omega n=1.5],scaledwidth=75.0% ] but for @xmath0 .
four curves correspond to sequences with the compactness @xmath65 , @xmath60 , @xmath62 , and @xmath64 . here
arrows denote the turning points where the angular momentum becomes minimum .
[ j - omega n=0.5 ] , scaledwidth=75.0% ] ( defined as eq .
( [ new omega ] ) ) is plotted against renormalized angular momentum @xmath86 ( defined as eq .
( [ new j ] ) ) for the sequences of constant rest mass binary systems with different values of the polytropic indices : @xmath0 ( dashed curves ) , @xmath1 ( solid curves ) , and @xmath2 ( dash - dotted curves ) .
numbers attached to each curve are the values of @xmath55 .
sequences of baumgarte _ et al . _
@xcite are also drawn by different symbols , as open circles for @xmath1 and @xmath104 , and @xmath62 , and crosses for @xmath2 and @xmath104 .
it should be noted that there exist no quasiequilibrium sequences with @xmath105 and @xmath97 for @xmath106 polytropes and with @xmath107 for @xmath108 polytropes because even the maximum mass configurations can not reach the corresponding values of @xmath55 for @xmath108 and @xmath109 polytropes .
[ j - omega new ] , scaledwidth=75.0% ] | we have computed quasiequilibrium sequences of synchronously rotating compact binary star systems with constant rest masses .
this computation has been carried out by using the numerical scheme which is different from the scheme based on the conformally flat assumption about the space .
stars are assumed to be polytropes with polytropic indices of @xmath0 , @xmath1 , and @xmath2 . since we have computed binary star sequences with a constant rest mass , they provide approximate evolutionary tracks of binary star systems . for
relatively stiff equations of state ( @xmath3 ) , there appear turning points along the quasiequilibrium sequences plotted in the angular momentum angular velocity plane .
consequently secular instability against exciting internal motion sets in at those points .
qualitatively , these results agree with those of baumgarte _ et al . _ who employed the conformally flat condition .
we further discuss the effect of different equations of state and different strength of gravity by introducing two kinds of dimensionless quantities which represent the angular momentum and the angular velocity .
strength of gravity is renormalized in these quantities so that the quantities are transformed to values around unity .
therefore we can clearly see relations among quasiequilibrium sequences for a wide variety of strength of gravity and for different compressibility . |
risk management is a main topic in the study of finance . in the present paper
, we consider the problem of minimizing the downside risk associated with an investor s total wealth in a certain incomplete market model .
more precisely , let @xmath2 be the price of a riskless asset with the dynamics @xmath3 , @xmath4 the prices of the risky assets , and @xmath5 , @xmath6 , the number of shares of @xmath7th security .
then the total wealth that the investor possesses is defined as @xmath8 and we assume a self - financing condition , @xmath9 when setting the proportion of the portfolio invested in the @xmath7th security as @xmath10 , we have @xmath11 and the total wealth is denoted by @xmath12 , which is the solution to this stochastic differential equation for a given strategy @xmath13 .
let us consider minimizing the probability @xmath14 for a given target growth rate @xmath15 by selecting portfolio choice @xmath16 .
let us make clear the meaning of the probability . if we choose strategy @xmath17 , then we have @xmath18 thus , the probability is always @xmath19 for large time @xmath0 and @xmath20 .
accordingly , in considering the above minimization , we investigate the extent for which we can improve the probability by selecting a strategy , as compared with the trivial strategy of investing the total wealth in a riskless asset . the latter strategy is considered the benchmark in terms of finance .
we shall consider the asymptotic behavior of the probability @xmath21 according to the theory of large deviation , it is natural to relate ( [ eq1.2 ] ) to @xmath22\ ] ] for @xmath23 .
namely , as @xmath1 , @xmath24 \biggr ) \to -\inf_{k\in ( -\infty,\kappa]}\sup_{\gamma<0}\{\gamma k-{\hat\chi}(\gamma)\}\ ] ] is expected to hold since the legendre transform @xmath25 of @xmath26 , @xmath27 is regarded as the rate function of the asymptotics , if @xmath26 is a convex function ; cf .
note that we can see from hlder s inequality that@xmath28=\log e[e^{\gamma\log({{v_t(h)}/{s_t^0}})}]$ ] is a convex function of @xmath29 , but this does not always imply the convexity of its infimum @xmath30 = \inf_{h_{\cdot}}\log e\bigl[e^{\gamma\log ( { { v_t(h)}/{s_t^0}})}\bigr].\ ] ] therefore , the convexity of @xmath31 can not be determined immediately and the above idea does not directly apply . in the present paper , we will find the convexity of @xmath32 by identifying the solution of the h - j - b equation of ergodic type with the limit value ( [ eq1.3 ] ) ; cf .
proposition [ prop4.2 ] and corollary [ cor4.1 ] .
then we shall see that the duality relation between ( [ eq1.2 ] ) and ( [ eq1.3 ] ) holds under suitable conditions , as expected ; cf .
theorem [ theo2.4 ] .
minimization ( [ eq1.4 ] ) , which is equivalent to power utility maximization , could be regarded as a risk - sensitive control problem .
the infinite time horizon counterpart of ( [ eq1.4 ] ) without a benchmark , @xmath33,\ ] ] has been extensively studied as risk - sensitive control ( e.g. , @xcite ) , and a benchmarked case has recently been reported in @xcite . from the viewpoint of stochastic control theory , it may appear more natural , compared with the above relationship between ( [ eq1.2 ] ) and ( [ eq1.3 ] ) , to relate @xmath34\ ] ] to @xmath35 which is considered in the present paper as well ; cf .
theorem [ theo2.5 ] .
we note that the problem relating ( [ eq1.2 ] ) to ( [ eq1.3 ] ) is thought to be equivalent to considering @xmath36 and @xmath37\ ] ] without a benchmark . however , the arguments in this article may be simpler than in the case without a benchmark ( cf .
remark [ rem2.2 ] ) . in previous papers @xcite , we studied similar asymptotic behavior without benchmarks for linear gaussian models in relation to the asymptotics of risk - sensitive portfolio optimization .
indeed , we established a duality relation between these problems , and as a result , an explicit expression of the limit value of the probability minimizing downside risk for each case of full and partial information . to obtain those results ,
the key analysis involved poisson equations derived by taking derivatives with respect to @xmath29 of the h - j - b equations of ergodic type corresponding to risk - sensitive control over an infinite time horizon .
since the solutions of the h - j - b equations can be explicitly expressed as quadratic functions by using the solutions of the riccati equations for linear gaussian models , analysis of the differentiability of the solutions of the riccati equations with respect to @xmath29 was essential in those works . in the present paper , we shall consider general diffusion market models and discuss the above - mentioned duality relation between the asymptotics of the minimization of downside risk and the risk - sensitive stochastic control for large time .
since the solutions of h - j - b equations of ergodic type do not always have explicit expressions , we need to consider , in general , the differentiablity with respect to @xmath29 of h - j - b equations of ergodic type .
the analysis is presented in sections [ sec5 ] and [ sec6 ] based on the results concerning h - j - b equations of ergodic type and related stochastic control problems given in sections [ sec3 ] and [ sec4 ] . here
, we mention the ongoing work of hata and sheu @xcite , which is closely related to the present paper and examines similar problems under the assumptions that @xmath38 in ( [ eq2.2 ] ) in section [ sec2 ] is bounded and that @xmath39 for @xmath40 in place of ( [ eq2.20 ] ) .
we shall explain more precisely the relationships between these papers in remark [ rem2.3 ] .
we note that maximization of an upside chance probability for the long term was studied by pham @xcite for continuous time models , and then by stettner @xcite for discrete time models , in relation to risk - sensitive portfolio optimization in the risk - seeking case . by regarding the maximization problem as large deviation control
, pham established a duality relation between these two types of problems . explicit calculation of the limit value is given in the case of 1-dimensional gaussian models .
the problem was later extended to a nonlinear case by hata and sekine @xcite and also to the partial information case by hata and iida @xcite for 1-dimensional gaussian models .
see also related works @xcite .
however , asymptotic estimates of downside risk probabilities and upside chance probabilities can not be obtained in parallel .
indeed , obtaining the estimates of downside risk is rather difficult than those of upside chance and further analysis of h - j - b equations is required to show the estimates as was shown in @xcite .
further note that large deviations control ( [ eq1.7 ] ) is an unconventional optimization problem , and thus we need to employ a new approach to study it .
consider a market model with @xmath41 securities and @xmath42 factors , where the bond price is governed by the ordinary differential equation @xmath43 the other security prices and factors are assumed to satisfy the stochastic differential equations @xmath44 \\[-8pt ] s^i(0)&=&s^i , \qquad i=1,\ldots , m , \nonumber\end{aligned}\ ] ] and @xmath45 \\[-8pt ] x(0)&=&x , \nonumber\end{aligned}\ ] ] where @xmath46 is a standard @xmath47-dimensional brownian motion process on a probability space @xmath48 .
let @xmath5 be the number of the shares of the @xmath7th security .
then , the total wealth that the investor possesses is defined as @xmath49 and the proportion of the portfolio invested in the @xmath7th security is @xmath50 @xmath51 [ or , @xmath52 is called self - financing if @xmath53 thus , under the self - financing condition , we have @xmath54 here we note that @xmath13 is defined as an @xmath55-vector consisting of @xmath56 since @xmath57 holds by definition .
the filtration that must be satisfied by admissible investment strategies @xmath58 is relevant in the present problem , and we introduce the following definition . @xmath59 is said to be an investment strategy if @xmath60 is an @xmath61 valued @xmath62-progressively measurable stochastic process such that @xmath63 the set of all investment strategies is denoted by @xmath64 . for a given @xmath65 ,
the process @xmath12 representing the total wealth of the investor at time @xmath66 is determined by the stochastic differential equation shown above .
@xmath67 \\[-8pt ] v_0&=&v_0 , \nonumber\end{aligned}\ ] ] where @xmath68 .
we are interested in the asymptotics of minimization of a downside risk for a given constant @xmath15 in comparison with investing the whole portfolio in a riskless security as the benchmark @xmath69 we also examine downside risk minimization with the benchmark @xmath70 over an infinite time horizon , @xmath71 where @xmath72 @xmath73 will be shown to be related to the following risk - sensitive asset allocation problem with benchmark @xmath70 .
namely , for a given constant @xmath74 , let us consider the asymptotics @xmath75 where @xmath76=\log e\bigl[e^{\gamma \log ( { { v_t(h)}/{s_t^0}})}\bigr ] , \ ] ] and @xmath16 ranges over the set @xmath77 of all admissible investment strategies defined by @xmath78=1\bigr\}.\ ] ] then we shall see that ( [ eq2.6 ] ) could be considered the dual problem to ( [ eq2.8 ] ) , while ( [ eq2.7 ] ) is the dual problem to risk - sensitive asset allocation over an infinite time horizon , @xmath79 where @xmath80 we shall consider these problems under the assumptions that @xmath81 and @xmath82 hold . in considering these problems , we first introduce the value function @xmath83.\ ] ] note that @xmath84 where @xmath85
. therefore @xmath86 where @xmath87 by introducing a probability measure @xmath88,\ ] ] the dynamics of the factor process can be written as @xmath89 with the new brownian motion process @xmath90 defined by @xmath91 and the value function written as @xmath92.\ ] ] the h - j - b equation for the value function @xmath93 is @xmath94+\frac{1}{2}(dv)^*\lambda\lambda^*dv\vspace*{1pt}\cr\displaystyle \qquad { } + \inf_h\{[\beta+\gamma \lambda\sigma^*h]^*dv+\gamma\eta(x , h)\}=0 , \vspace*{2pt}\cr\displaystyle v(t , x)=\gamma\log v_0 , } \ ] ] which is also written as @xmath95+\beta_{\gamma}^*dv+\frac{1}{2}(dv)^*\lambda n_{\gamma } ^{-1}\lambda^*dv - u_{\gamma}=0 , \vspace*{2pt}\cr\displaystyle v(t , x)=\gamma\log v_0 , } \ ] ] where @xmath96 and @xmath97 @xmath98^*dv+\gamma\eta(x , h)\ } \\ & & \qquad = \inf_{h\in r^m}\biggl\{[\gamma\lambda\sigma^*h]^*dv-\frac{\gamma ( 1-\gamma)}{2}h^*\sigma\sigma^*h+\gamma h^*{\hat\alpha}\biggr\ } \\ & & \qquad = \inf_{h\in r^m}\biggl\{-\frac{\gamma(1-\gamma)}{2}\biggl[h-\frac { 1}{1-\gamma}(\sigma\sigma^*)^{-1}({\hat\alpha}+\sigma\lambda ^*dv)\biggr]^*\\ & & \qquad\hphantom{=\inf_{h\in r^m}\biggl\ { } { } \times\sigma\sigma^*\biggl[h-\frac{1}{1-\gamma}(\sigma\sigma ^*)^{-1}({\hat\alpha}+\sigma\lambda^*dv)\biggr ] \\ & & \qquad\hphantom{=\inf_{h\in r^m}\biggl\ { } { } + \frac{\gamma}{2(1-\gamma)}({\hat\alpha}+\sigma\lambda ^*dv)^*(\sigma\sigma^*)^{-1}({\hat\alpha}+\sigma\lambda^*dv)\biggr\}. \ ] ] therefore , the function @xmath99 defines the generator of the optimal diffusion @xmath100 for @xmath101 : @xmath102+\biggl[\beta+\frac{\gamma}{1-\gamma}\lambda\sigma^*(\sigma\sigma ^*)^{-1}({\hat\alpha}+\sigma\lambda^*dv)\biggr]^*d\psi,\ ] ] which is seen in proposition [ prop2.1 ] .
set @xmath103 then , @xmath104+\beta_{\gamma}^*d{\bar v}-\frac{1}{2}(d{\bar v})^*\lambda n_{\gamma}^{-1}\lambda^*d{\bar v}+u_{\gamma}=0 , \vspace*{2pt}\cr\displaystyle { \bar v}(t , x)=-\gamma\log v_0 . } \ ] ] since @xmath105 , which is easily seen by taking @xmath106 with @xmath107 orthogonal to the range of @xmath108 , and seeing that @xmath109 , we have @xmath110 as for existence of a solution to ( [ eq2.16 ] ) satisfying sufficient regularities , we have the following results ; cf . @xcite .
[ theo2.1 ] assume ( [ eq2.11 ] ) and ( [ eq2.12 ] )
. then h - j - b equation ( [ eq2.16 ] ) has a solution such that @xmath111 { \bar v } , \frac{\partial{\bar v}}{\partial t } , \frac{\partial { \bar v}}{\partial x_k } , \frac{\partial^2 { \bar v}}{\partial x_k\,\partial x_j}&\in & l^p(0,t;l^p_{\mathrm{loc}}(r^n ) ) , \qquad 1<\forall p<\infty , \\[-2pt ] \frac{\partial^2{\bar v}}{\partial t^2 } , \frac{\partial^2{\bar v}}{\partial x_k\,\partial t } , \frac{\partial^3 { \bar v}}{\partial x_k\,\partial x_j\,\partial x_l } , \frac{\partial^3 { \bar v}}{\partial x_k\,\partial x_j\,\partial t}&\in & l^p(0,t;l^p_{\mathrm{loc}}(r^n ) ) , \\[-2pt ] \frac{\partial{\bar v}}{\partial t}&\leq&0,\end{aligned}\ ] ] and @xmath112 & & \hphantom{c\bigl ( } { } + |\nabla\beta_{\gamma}|_{2\rho}+|\beta_{\gamma } |_{2\rho}^2+|u_{\gamma}|_{2\rho}+|\nabla u_{\gamma}|_{2\rho}+1\bigr ) \ ] ] for @xmath113 and @xmath114 , where @xmath115 , @xmath116 , @xmath117 , @xmath118 is a universal constant and @xmath119 . for @xmath120 , we consider the stochastic differential equation @xmath121 and define @xmath122 for the solution @xmath123 of the stochastic differential equation . note that the solution of this stochastic differential equation is obtained by the change of measure from the solution of ( [ eq2.3 ] ) .
indeed , we can see that @xmath124 has at most linear growth under assumptions ( [ eq2.11 ] ) and ( [ eq2.12 ] ) from the above gradient estimates , and therefore , @xmath125=1\ ] ] holds .
thus @xmath126\vadjust{\goodbreak}\ ] ] defines a probability measure . under this measure
, @xmath123 turns out to be a solution of the above stochastic differential equation .
the following is a verification theorem , the proof of which is almost the same as the proof of proposition 2.1 , @xcite , and thus is omitted here .
[ prop2.1 ] assume ( [ eq2.11 ] ) and ( [ eq2.12 ] ) .
then @xmath127 and it is optimal @xmath128=\log e\bigl[e^{\gamma(\log v_t({\hat h})-\log s_t^0)}\bigr].\ ] ] let us consider an h - j - b equation of ergodic type that is thought to be the limit equation of ( [ eq2.15 ] ) .
namely , @xmath129+\beta_{\gamma } ^*d w+\tfrac{1}{2}(d w)^*\lambda n_{\gamma}^{-1}\lambda^*d w - u_{\gamma}.\ ] ] set @xmath130 and assume that @xmath131 and @xmath132 under these assumptions , we have a solution to the h - j - b equation of ergodic type , and the proof is given in proposition [ prop3.1 ] in section [ sec3 ] .
assume ( [ eq2.11 ] ) , ( [ eq2.12 ] ) , ( [ eq2.20 ] ) and ( [ eq2.21 ] ) .
then ( [ eq2.19 ] ) has a solution @xmath133 such that @xmath134 , @xmath135 and the solution satisfying this condition is unique up to additive constants with respect to @xmath136 .
we further assume that @xmath137 then we have the following theorem , and the proof is given after proposition [ prop4.2 ] in section [ sec4 ] .
[ theo2.2 ] under assumptions ( [ eq2.11 ] ) , ( [ eq2.12 ] ) , ( [ eq2.20 ] ) and ( [ eq2.22 ] ) , we have @xmath138 the following results are important to prove our main results , and the proofs are given in lemma [ lem6.3 ] , lemma [ lem7.1 ] and corollary [ cor4.1 ] .
[ theo2.3 ] let @xmath139 be a solution to ( [ eq2.19 ] )
. then under the assumptions of theorem [ theo2.2 ] , @xmath140 and @xmath141 are differentiable with respect to @xmath29 and @xmath140 is convex .
their derivatives satisfy @xmath142+\bigl(\beta_{\gamma}^*+\bigl(d w^{(\gamma)}\bigr)^*\lambda n_{\gamma } ^{-1}\lambda^*\bigr)d w_{\gamma } \nonumber \\[-8pt ] \\[-8pt ] & & { } + \frac{1}{2(1-\gamma)^2}\bigl\{{\hat\alpha}+\sigma \lambda^*dw^{(\gamma)}\bigr\}^*(\sigma\sigma^*)^{-1}\bigl\{{\hat\alpha } + \sigma\lambda^*dw^{(\gamma)}\bigr\ } , \nonumber\end{aligned}\ ] ] where @xmath143 .
furthermore , @xmath144 [ rem2.2 ] it is important to know the limit value @xmath145 since it determines the left endpoint of the interval of the target growth rate @xmath15 , which makes @xmath146 finite .
here we compare the results above with those to be expected for the case without a benchmark , considering asymptotics @xmath37.\ ] ] the h - j - b equation of ergodic type of this problem becomes @xmath147+\beta_{\gamma}^*d{\check w}+\tfrac{1}{2}(d { \check w})^*\lambda n_{\gamma}^{-1}\lambda^*d { \check w}-u_{\gamma}+\gamma r(x),\ ] ] and we can obtain its derivative @xmath148+\bigl(\beta_{\gamma}^*+(d { \check w})^*\lambda n_{\gamma}^{-1}\lambda^*\bigr)d { \check w}_{\gamma } \\ & & { } + \frac{1}{2(1-\gamma)^2}\{{\hat\alpha}+\sigma \lambda^*d{\check w}\}^*(\sigma\sigma^*)^{-1}\{{\hat\alpha}+\sigma \lambda^*d{\check w}\}+r \ ] ] through almost the same arguments as the current ones provided to obtain the results in the present article
. the difference appears in considering the asymptotics of @xmath149 as @xmath150 .
indeed , @xmath151 could be seen as in @xcite , where @xmath152 is the invariant measure of @xmath153-diffusion process and @xmath154 is defined by @xmath155+(\beta_{\gamma}+\lambda n_{\gamma}^{-1}\lambda^*d{\check w})^*d\psi.\ ] ] note that @xmath153 corresponds to @xmath156 defined by ( [ eq4.16 ] ) in the present paper and can be shown to be ergodic under suitable conditions in a manner similar to the proof of proposition [ prop4.3 ] .
now we can state our main theorems .
the proofs are given in sections [ sec7 ] and [ sec8 ] .
[ theo2.4 ] under the assumptions of theorem [ theo2.2 ] , for @xmath157 , @xmath158}\sup_{\gamma<0}\{\gamma k-{\hat\chi}(\gamma)\}=-\sup_{\gamma<0}\{\gamma\kappa-{\hat\chi } ( \gamma)\}.\ ] ] moreover , for @xmath159 such that @xmath160 , take a strategy @xmath161 defined in proposition [ prop2.1 ] . then , @xmath162 for @xmath163 , @xmath164 for the solution @xmath165 to h - j - b equation ergodic type ( [ eq2.19 ] ) , let us set @xmath166 further consider the stochastic differential equation @xmath167 and define @xmath168 for the solution @xmath123 of the stochastic differential equation .
then we have the following theorem .
[ theo2.5 ] under the assumptions of theorem [ theo2.2 ] , let @xmath157 and @xmath159 be the same as above .
we also assume that @xmath169 then @xmath170}\sup _ { \gamma<0}\{\gamma k-{\hat\chi}(\gamma)\}=-\sup_{\gamma<0}\ { \gamma\kappa-{\hat\chi}(\gamma)\}\ ] ] and @xmath171 [ rem2.3 ] in our previous paper @xcite , we studied similar problems without benchmarks in the case of linear gaussian models . specifically , we discussed the case where @xmath172 , @xmath173 , @xmath174 , @xmath175 and @xmath176 , in which @xmath177 , @xmath178 , @xmath179 and @xmath180 ( resp . ,
@xmath181 and @xmath182 ) are all constant matrices ( resp . , vectors ) , and @xmath183 is a constant . under the main assumption that @xmath184 which corresponds to ( [ eq2.20 ] ) above , we obtained results similar to theorems [ theo2.4 ] and [ theo2.5 ] . the present paper is a natural extension to general diffusion incomplete market models . on the other hand , hata and sheu
@xcite treat the case where @xmath38 is bounded , and @xmath39 for @xmath185 , in which linear gaussian models are excluded . in that case
, @xmath186 becomes bounded and they employ quite different methods from ours to analyze h - j - b equation ( [ eq2.19 ] ) , while assumption ( [ eq2.22 ] ) is crucial in our settings .
for that reason our theorems do not include the case where @xmath38 is bounded .
the generator of the optimal diffusion process governed by ( [ eq2.25 ] ) for risk - sensitive control problem ( [ eq2.10 ] ) is defined by @xmath187+\biggl[\beta_{\gamma}^*+\frac{\gamma}{1-\gamma}(dw)^*\lambda\sigma ^*(\sigma\sigma^*)^{-1}\sigma\lambda^*\biggr]d\psi.\ ] ] on the other hand , in proving theorem [ theo2.2 ] we introduce another type of stochastic control problem ( [ eq4.9 ] ) with ( [ eq4.7 ] ) .
the generator of the optimal diffusion process for this problem is defined by ( [ eq4.16 ] ) .
@xmath188+[\beta_{\gamma}^*+(dw)^*\lambda n_{\gamma}^{-1}\lambda^*]d\psi,\ ] ] where @xmath136 is a solution to h - j - b equation ( [ eq2.19 ] ) of ergodic type
. then we note that @xmath156 is related to @xmath189 through the gauge transform , @xmath190\varphi=\bigl[{\bar l}-\bigl(\gamma\eta-\chi(\gamma ) \bigr)\bigr]\varphi.\ ] ] further , we see that @xmath191 is an eigenfunction of @xmath192 @xmath193 for the principal eigenvalue @xmath140 ; cf .
note that @xmath156 is ergodic as is seen in proposition [ prop4.3 ] , while @xmath189 is not always ergodic .
we assume ( [ eq2.11 ] ) and ( [ eq2.12 ] ) and that @xmath194 , @xmath195 with an @xmath196 ( resp . , @xmath197 ) matrix - valued bounded function @xmath177 ( resp . , @xmath178 ) , and
an @xmath55 ( resp . , @xmath42)-vector - valued bounded function @xmath181 ( resp . ,
@xmath182 ) such that : @xmath198 the real parts of all eigenvalues of @xmath199 is less than @xmath200 , @xmath201 @xmath202 . in this case
@xmath203 and we see that ( [ eq2.20 ] ) holds . furthermore , ( [ eq2.22 ] ) holds because of ( i ) .
instead of ( [ eq2.19 ] ) , we shall study an h - j - b equation of ergodic type for @xmath204 . @xmath205+\beta _ { \gamma}^*d{\bar w}-\tfrac{1}{2}(d{\bar w})^*\lambda n_{\gamma } ^{-1}\lambda^*d{\bar w}+u_{\gamma}.\ ] ] [ prop3.1 ] assume ( [ eq2.11 ] ) , ( [ eq2.12 ] ) , ( [ eq2.20 ] ) and ( [ eq2.21 ] ) .
then ( [ eq3.1 ] ) has a solution @xmath206 such that @xmath207 , @xmath208 and the solution satisfying this condition is unique up to additive constants with respect to @xmath209 .
the following notation is useful for the task at hand .
set @xmath210 .
then @xmath211 moreover , we see that @xmath212 to prove proposition [ prop3.1 ] , we first consider the h - j - b equation of discounted type , @xmath213+\beta_{\gamma}^*d v_{{\varepsilon}}-\tfrac{1}{2}(d v_{{\varepsilon}})^*\lambda n_{\gamma}^{-1}\lambda^*d v_{{\varepsilon}}+u_{\gamma}.\ ] ] note that ( [ eq3.2 ] ) can be written as @xmath214+g^*d v_{{\varepsilon}}-\tfrac{1}{2}(\lambda^ * d v_{{\varepsilon}}-\sigma^*{\hat\alpha})^ * n_{\gamma}^{-1}(\lambda^*d v_{{\varepsilon}}-\sigma^*{\hat\alpha})\nonumber\hspace*{-15pt } \\[-9pt ] \\[-9pt ] & & { } + \tfrac{1}{2}{\hat\alpha}\sigma\sigma ^*{\hat\alpha}. \nonumber\end{aligned}\ ] ] then , we consider the linear equation @xmath215 where @xmath216+g^*d\varphi.\ ] ] under assumptions ( [ eq2.11 ] ) , ( [ eq2.12 ] ) and ( [ eq2.20 ] ) , ( [ eq3.4 ] ) has a solution @xmath217 . indeed , set @xmath218 , @xmath219 .
then , by taking @xmath220 to be sufficiently large , we can see that there exists @xmath221 such that for @xmath222 , @xmath223 therefore , when setting @xmath224 @xmath225 turns out to be a supersolution to ( [ eq3.4 ] ) , and we can see that there exists a solution @xmath217 to ( [ eq3.4 ] ) such that @xmath226 since @xmath227 is a subsolution .
we note that @xmath228 is a supersolution to ( [ eq3.2 ] ) which is the same equation as ( [ eq3.3 ] ) .
[ lem3.1 ] under the assumptions of proposition [ prop3.1 ] , ( [ eq3.2 ] ) has a solution such that @xmath229 and @xmath230 . in proving the existence of the solution
, we introduce a dirichlet problem on @xmath231 , @xmath232 : @xmath233+\beta_{\gamma}^*d v_{{\varepsilon}}-\tfrac{1}{2}(d v_{{\varepsilon}})^*\lambda n_{\gamma}^{-1}\lambda^*d v_{{\varepsilon}}+u_{\gamma } \quad \mbox{in } b_r , \nonumber\hspace*{-15pt } \\[-9pt ] \\[-9pt ] \qquad v_{{\varepsilon}}(x)&=&\varphi_{{\varepsilon } } ,
\qquad x\in \partial b_r . \nonumber\end{aligned}\ ] ] owing to theorem 8.3 ( @xcite , chapter 4 ) , dirichlet problem ( [ eq3.4 ] ) has a solution @xmath234 .
we extend @xmath234 to the whole euclidean space as @xmath235 then we can see that @xmath236 is nonincreasing with respect to @xmath237 . indeed , for @xmath238 , @xmath236 is a supersolution to ( [ eq3.3 ] ) in @xmath239 , and we have @xmath240+\beta _ { \gamma}^*d(v_{{\varepsilon},r}-v_{{\varepsilon},r ' } ) \\ & & \qquad \quad { } -\tfrac{1}{2}(d v_{{\varepsilon},r})^*\lambda n_{\gamma}^{-1}\lambda^*d v_{{\varepsilon},r}+ \tfrac{1}{2}(d v_{{\varepsilon},r'})^*\lambda n_{\gamma}^{-1}\lambda^*d v_{{\varepsilon},r ' } \qquad \mbox{in } b_r ' \\ & & \qquad = \tfrac{1}{2}\operatorname{tr}[\lambda\lambda^*d^2 ( v_{{\varepsilon},r}-v_{{\varepsilon},r'})]+\beta_{\gamma}^*d ( v_{{\varepsilon},r}-v_{{\varepsilon},r ' } ) \\ & & \qquad \quad { } -\tfrac{1}{2}(d v_{{\varepsilon},r}+d v_{{\varepsilon},r'})^*\lambda n_{\gamma } ^{-1}\lambda^*d ( v_{{\varepsilon},r}-v_{{\varepsilon},r ' } ) . \ ] ] therefore , from the maximum principle ( cf .
theorem 3.1 in @xcite ) we see that @xmath241 since @xmath242 .
we further note that @xmath243 for each @xmath237 because @xmath244 is a subsolution to ( [ eq3.2 ] ) , and the maximum principle again applies .
similar to the proof of lemma 2.6 in @xcite , we have the following gradient estimate : for each @xmath237 and @xmath245 , @xmath246 where @xmath247 is a constant independent of @xmath237 , @xmath248 .
thus , when taking a sequence @xmath249 such that @xmath250 , @xmath251 forms a family of uniformly bounded and equicontinuous functions .
thus we can choose a subsequence @xmath252 converging to a continuous function @xmath234 .
furthermore , since @xmath253 for a positive constant @xmath254 independent of @xmath249 and @xmath248 , it converges weakly in @xmath255 to @xmath234 by taking a subsequence if necessary . by similar arguments to lemma 6.8 in @xcite
, the convergence can be strengthened as @xmath256 converges strongly in @xmath257 to @xmath258 . as a result
we can see from the regularity theorems that we have a solution @xmath229 to ( [ eq3.2 ] ) .
since @xmath259 , for each @xmath232 from the maximum principle as well as ( [ eq3.7 ] ) , we see that @xmath260 .
set @xmath261 then , by taking @xmath262 to be sufficiently small , we can see that there exists @xmath263 such that for @xmath264 , @xmath265 therefore , we see that @xmath266 and @xmath267 satisfy assumption ( [ eq9.3 ] ) in the last section .
set @xmath268 , @xmath269 and @xmath270 then for @xmath271 there exists a solution @xmath272 to @xmath273 if and only if @xmath274 where @xmath275 is an invariant measure for @xmath266 ; cf . proposition [ prop9.4 ] in in the .
therefore , setting @xmath276 there exists a solution @xmath277 to @xmath278 and it is known that @xmath279 converges to @xmath280 as @xmath281 uniformly on each compact set .
now we can prove proposition [ prop3.1 ] .
proof of proposition [ prop3.1 ] we first note that @xmath282 because of lemma [ lem3.1 ] .
therefore , we have @xmath283 where @xmath284 is a constant independent of @xmath248 .
moreover , @xmath285 for a positive constant @xmath286 independent of @xmath248 in view of ( [ eq3.8 ] ) .
thus , similarly to the proof of theorem 3.1 in @xcite , we can prove the existence of the solution @xmath287 to ( [ eq3.1 ] ) such that @xmath288 . from regularity theorems
we see that @xmath207 .
the proof of uniqueness is similar to the proof of lemma 3.2 in @xcite .
now we have the following proposition . under the assumptions of proposition [ prop3.1 ] , the solution @xmath209 to ( [ eq3.1 ] ) satisfies @xmath289 ) ,
then , for each @xmath290 , there exists a positive constant @xmath291 such that the nonnegative solution @xmath292 , @xmath293 , satisfies @xmath294 set @xmath295
. then we shall prove for each @xmath296 that @xmath297 \\[-8pt ] & & \hphantom{k\biggl(}{}+|u_{\gamma } |_r+|\nabla u_{\gamma}|_r+\frac{|\beta_{\gamma}|_r}{r}+|\nabla \beta_{\gamma}|_r+c\biggr ) \nonumber\end{aligned}\ ] ] for positive constants @xmath298 and @xmath220 , where @xmath299 .
note that ( [ eq3.13 ] ) implies ( [ eq3.11 ] ) because of our assumptions on the coefficients @xmath300 and @xmath183 .
we have @xmath301 since @xmath302 and @xmath303 as @xmath281 . in the following @xmath304 , @xmath305 and @xmath186
are abbreviated to @xmath306 , @xmath307 and @xmath308 , respectively .
@xmath309 is abbreviated to @xmath310 . by differentiating ( [ eq3.1 ] ) with respect to @xmath311
, we have @xmath312 \\[-8pt ] & & { } -d_iwq^{ij}d_{jk}w-\tfrac{1}{2}d_iwq^{ij}_kd_jw+u_k . \nonumber \ ] ] set @xmath313 then we have @xmath314 here we have used ( [ eq3.14 ] ) and the matrix inequality @xmath315)^2\leq n c\operatorname{tr}[ab^2]\]]for symmetric matrix @xmath178 and nonnegative definite symmetric matrix @xmath177 , where @xmath118 is the maximum eigenvalue of @xmath177 .
set @xmath316 then @xmath317\geq-\frac{4n}{r^2}c_2 , $ ] @xmath318 and @xmath319 .
let @xmath320 be the maximum point of @xmath321 in @xmath322 .
then @xmath323 and @xmath324(x)\leq0 $ ] . therefore ,
from the maximum principle we have @xmath325 \\ & & { } -f\biggl\{\frac{1}{2}(\lambda\lambda^*)^{ij}d_{ij}\tau-\frac{(\lambda \lambda^*)^{ij}d_i\tau d_j\tau}{\tau } -\beta^id_i\tau + q^{ij}d_jwd_i\tau\biggr\}. \ ] ] since @xmath326 , by taking @xmath262 to be sufficiently small , @xmath327 therefore , it follows that @xmath328 we can assume @xmath329 and @xmath330 ; thus , @xmath331 accordingly , we have @xmath332 therefore , we obtain @xmath333 and universal constant @xmath219 . including the case where @xmath334 , @xmath335 , we obtain @xmath336 and ( [ eq3.13 ] ) has been proved .
now let us prove ( [ eq3.12 ] ) . for each @xmath337 take a point @xmath338 such that @xmath339 .
set @xmath340 where @xmath341 is a positive constant determined later
. then , @xmath342 in @xmath343 and @xmath344 on @xmath345 .
set @xmath346 then , @xmath347 note that @xmath348 then we have @xmath349-\beta ^*dz \\ & & \qquad
= -\frac{1}{2}(d{\bar w})^*\lambda n_{\gamma}^{-1}\lambda^*d{\bar w}+u_{\gamma}+\frac{1}{2}\operatorname{tr}[\lambda\lambda^*d^2r]+\beta ^*dr \\ & & \qquad = -\frac{1}{2}d({\bar w}+r)^*\lambda n_{\gamma}^{-1}\lambda ^*d({\bar w}-r)-\frac{1}{2}(dr)^*\lambda n_{\gamma}^{-1}\lambda ^*dr+u_{\gamma } \\ & & \qquad \quad { } + \frac{1}{2}\operatorname{tr}[\lambda\lambda^*d^2r]+\beta^*dr \\ & & \qquad \geq -\frac{1}{2}d({\bar w}+r)^*\lambda n_{\gamma}^{-1}\lambda^*dz-\frac { c_2}{2}|dr|^2+u_{\gamma}+\frac{1}{2}\operatorname{tr}[\lambda\lambda ^*d^2r]+\beta^*dr . \ ] ] noting that @xmath350 , for a positive constant independent of @xmath29 , @xmath349-\beta
^*dz+\frac{1}{2}d({\bar w}+r)^*\lambda n_{\gamma}^{-1}\lambda^*dz \\ & & \qquad \geq-\frac{c_2}{2}|dr|^2+u(x)-\frac{4c_{\rho}}{\rho^2}\operatorname { tr}[\lambda\lambda^*]-4c c{\rho } \\ & & \qquad \geq -\frac{8c_2c_{\rho}^2}{\rho^2}+\frac{-\gamma}{2(1-\gamma ) } c_0\biggl(\frac{|\rho|^2}{4}+1\biggr)-\frac{4c_2nc_{\rho}}{\rho^2}-4cc_{\rho } \\ & & \qquad \geq-\biggl(8c_2\frac{c_{\rho}^2}{\rho^2}+4c_2n\frac{c_{\rho}}{\rho ^2}+4cc_{\rho}\biggr)+\frac{-\gamma_0c_0\rho^2}{8(1-\gamma_0)}+\frac { -\gamma_0c_0}{2(1-\gamma_0)}. \ ] ] by setting @xmath351 with @xmath352 such that @xmath353 and @xmath354 , we see that @xmath355-\beta^*dz+\tfrac { 1}{2}d({\bar w}+r)^*\lambda n_{\gamma}^{-1}\lambda^*dz\geq m>0\qquad \mbox{in } d_{\rho}\ ] ] for some positive constant and sufficiently large @xmath356 .
then @xmath357 is superharmonic in @xmath343 and @xmath358 .
therefore @xmath359 , from which we have @xmath360 hence , @xmath361 .
let us come back to h - j - b equation ( [ eq2.16 ] ) .
according to assumption ( [ eq2.11 ] ) , we have a positive constant @xmath362 such that @xmath363 we strengthen condition ( [ eq2.21 ] ) to ( [ eq2.22 ] )
. then we have the following lemma .
[ lem4.1]assume ( [ eq2.11 ] ) , ( [ eq2.12 ] ) , ( [ eq2.22 ] ) and @xmath364 .
then for each @xmath365 there exist positive constants @xmath366 and @xmath367 such that @xmath368 choose a positive constant @xmath220 such that @xmath369 and set @xmath370 , where @xmath371 , and set @xmath372 where @xmath373 is a solution to the riccati equation @xmath374 and @xmath375 is a solution to the ordinary equation @xmath376-\frac{c c_{\beta } } { 2}-c_{\gamma}'=0,\qquad q(t)=-\gamma\log v_0,\ ] ] where @xmath377 .
set @xmath378 then @xmath379-\beta_{\gamma}^*dz \\ & & \qquad = \frac{1}{2}(d{\bar v})^*\lambda n_{\gamma}^{-1}\lambda^*d{\bar v}+u_{\gamma}+\frac{\partial r}{\partial t}+\frac{1}{2}\operatorname { tr}[\lambda\lambda^*d^2r]+\beta^*dr \\ & & \qquad = -\frac{1}{2}d({\bar v}+r)^*\lambda n_{\gamma}^{-1}\lambda^ * d({\bar v}-r)-\frac{1}{2}dr^*\lambda n_{\gamma}^{-1}\lambda^ * dr+u_{\gamma } \\ & & \qquad \quad { } + \frac{\partial r}{\partial t}+\frac{1}{2}\operatorname{tr}[\lambda \lambda^*d^2r]+\beta^*dr \\ & & \qquad \geq-\frac{1}{2}d({\bar v}+r)^*\lambda n_{\gamma}^{-1}\lambda^ * d({\bar v}-r)-\frac{c_2}{2(1-\gamma)}(dr)^*i_ndr+c_{\gamma } |x|^2-c_{\gamma } ' \\ & & \qquad \quad { } + \frac{1}{2}x^*{\dot p}(t)x+{\dot q}(t)+\frac{c_1}{2}\operatorname { tr}[p(t)]-\frac{c}{2}\beta_{\gamma}^*\beta_{\gamma}-\frac{1}{2c}(dr)^*dr . \ ] ] therefore , @xmath380-\beta^*dz+\frac{1}{2}d({\bar v}+r)^*\lambda n_{\gamma}^{-1}\lambda^*dz \\ & & \qquad \geq\frac{1}{2}x^*{\dot p}(t)x-\frac{1}{2}\biggl(\frac{c_2}{1-\gamma } + \frac{1}{c}\biggr)x^*p(t)i_np(t)x+\biggl(c_{\gamma}-\frac{cc_\beta } { 2}\biggr)|x|^2 \\ & & \qquad \quad { } + { \dot q}(t)+\frac{c_1}{2}\operatorname{tr}[p(t)]-\frac { cc_{\beta}}{2}-c_{\gamma } ' \\ & & \qquad \geq\frac{1}{2}\biggl ( c_{\gamma}-\frac{cc_\beta}{2}\biggr)|x|^2\geq0 . \ ] ] thus we see that @xmath381 is super harmonic in @xmath382 , and @xmath383
. therefore we have @xmath384 , that is , @xmath385 since @xmath373 is positive definite , @xmath386 let us rewrite ( [ eq2.16 ] ) as @xmath387+g^*d{\bar v } \cr\displaystyle
\hphantom{0=}{}-\frac{1}{2}(\lambda ^*d{\bar v}-\sigma^*{\hat\alpha})^ * n_{\gamma}^{-1}(\lambda ^*d{\bar v}-\sigma^*{\hat\alpha})+\frac{1}{2}{\hat\alpha}^*\sigma\sigma^*{\hat \alpha } , \vspace*{4pt}\cr\displaystyle { \bar v}(t , x)=-\gamma\log v_0 . }
\ ] ] noting that @xmath388 , \ ] ] we can rewrite it again as @xmath389+g^*d{\bar v}+\inf_{z\in r^{n+m}}\ { ( \lambda z)^*d{\bar v}+\varphi(x , z)\},\cr\displaystyle { \bar v}(t , x)=-\gamma\log v_0 , } \ ] ] where @xmath390 this h - j - b equation corresponds to the following stochastic control problem , the value of which is defined as @xmath391,\ ] ] where @xmath392 is a controlled process governed by the stochastic differential equation @xmath393 with control @xmath394 . here , @xmath395 is the set of all @xmath396 valued progressively measurable processes such that @xmath397<\infty.\ ] ] to study this problem , we introduce a value function for @xmath398 , @xmath399.\ ] ] by the verification theorem , the solution @xmath400 to ( [ eq4.5 ] ) can be identified with the value function @xmath401 .
indeed , set @xmath402 which attains the infimum in ( [ eq4.5 ] ) , and consider the stochastic differential equation @xmath403 from the estimates obtained in theorem [ theo2.1 ] , we see that ( [ eq4.8 ] ) has a unique solution .
it is also seen by using it s theorem that @xmath404\ ] ] holds , where @xmath405 . in a similar way
, we can see that @xmath406\ ] ] for each @xmath407 , hence , @xmath408 .
let us consider the following stochastic control problem with the averaging cost criterion : @xmath409,\ ] ] where @xmath392 is a controlled process governed by controlled stochastic differential equation ( [ eq4.7 ] ) with control @xmath410 .
the solution @xmath392 of ( [ eq4.7 ] ) is sometimes written as @xmath411 to make clear the dependence on control @xmath410 , and the set @xmath412 of all admissible controls is defined as follows : @xmath413=0 , e\biggl[\int_0^t|z_s|^2\,ds\biggr]<\infty , \forall t\biggr\}. \ ] ] corresponding to this stochastic control problem , h - j - b equation of ergodic type ( [ eq3.1 ] ) can be written as @xmath414+g^*d{\bar w}+\inf_{z\in r^{n+m}}\{(\lambda z)^*d{\bar w}+\varphi
( x , z)\}.\ ] ] we then set @xmath415 and consider stochastic differential equation @xmath416 we shall prove the following proposition .
[ prop4.1 ] @xmath417 and @xmath418,\ ] ] where @xmath419 . for the proof of this proposition ,
we prepare the following lemma .
[ lem4.2 ] under assumptions ( [ eq2.11 ] ) , ( [ eq2.12 ] ) , ( [ eq2.20 ] ) and ( [ eq2.22 ] ) the following estimates hold . for each @xmath420 there exist positive constants @xmath421 and
@xmath422 independent of @xmath0 and @xmath29 with @xmath423 such that @xmath424\leq c\ ] ] and also @xmath425\leq c.\ ] ] let us set @xmath426+(g+\lambda{\hat z})^*d\psi \nonumber \\[-8pt ] \\[-8pt ] & = & \tfrac{1}{2}\operatorname{tr}[\lambda\lambda^*d^2\psi]+(\beta_{\gamma } -\lambda n_{\gamma}^{-1}\lambda^*d{\bar w})^*d\psi.\nonumber \ ] ] then we have @xmath427 therefore , by applying it s formula , we have @xmath428 thus , for @xmath429 , @xmath430 taking into account ( [ eq2.22 ] ) and ( [ eq2.17 ] ) , for @xmath421 such that @xmath431 , we have @xmath432 for @xmath433 .
thus , we obtain @xmath434 therefore , taking @xmath435 and setting @xmath436 we see that @xmath437&\leq & e^{\delta{\bar w}(x)}+\delta e\biggl[\int_0^{t\wedge\tau } e^{p\delta s+\delta{\bar w}({\bar y}_s)}\{-k_1|{\bar y}_s|^2+k_2\}\,ds\biggr ] \\ & \leq & e^{\delta{\bar w}(x)}+\delta k_2e\biggl[\int_0^{t\wedge \tau}e^{p\delta s+\delta{\bar w}({\bar y}_s)}{\mathbf1}_{\{|{\bar y}_s|^2\leq{{k_2}/{k_1}}\}}\,ds\biggr ] \\ & \leq & e^{\delta{\bar w}(x)}+\delta k_2e^{\delta k_3}e\biggl[\int_0^{t\wedge\tau}e^{p\delta s}\,ds\biggr ] \\ & = & e^{\delta{\bar w}(x)}+ k_2e^{\delta k_3 } e\biggl[\frac { 1}{p}\bigl(e^{p\delta(t\wedge\tau)}-1\bigr)\biggr ] . \ ] ] by letting @xmath237 tend to @xmath438 , we have @xmath439\leq e^{\delta{\bar w}(x)}+ k_2e^{\delta k_3}\frac{1}{p}(e^{p\delta t}-1).\ ] ] hence , @xmath440&\leq & e^{-p\delta t+\delta{\bar w}(x)}+ k_2e^{\delta k_3}\frac{1}{p}(1-e^{-p\delta t } ) \\ & \leq & e^{\delta{\bar w}(x)}+ k_2e^{\delta k_3}\frac{1}{p}. \ ] ] finally , we see that ( [ eq4.15 ] ) follows from ( [ eq4.14 ] ) because of ( [ eq3.12 ] ) .
proof of proposition [ prop4.1 ] from ( [ eq4.10 ] ) it follows that @xmath441+g^*d{\bar w}+(\lambda z)^*d{\bar w}+\varphi(x , z)\vadjust{\goodbreak}\ ] ] for each @xmath442 .
therefore , for each control @xmath410 we have @xmath443(y_s)\,ds \\ & \geq&\int_0^t(d{\bar w}(y_s))^*\lambda(y_s)\,dw_s-\chi t-\int _ 0^t\varphi(y_s , z_s)\,ds . \ ] ] thus , @xmath444,\ ] ] from which we obtain @xmath445 \\ & = & \limsup_{t\to\infty}\frac{1}{t}e\biggl[\int_0^t\varphi(y_s , z_s)\,ds\biggr ] \ ] ] since @xmath446 .
namely , we have @xmath447 . on the other hand , by taking @xmath448 , we have @xmath449 and thus @xmath450.\ ] ] lemma [ lem4.2 ] implies @xmath451\geq\rho,\ ] ] and we see that @xmath417 . let us define @xmath452=\limsup_{t\to \infty}\frac{1}{t}{\bar v}(0,x;t).\ ] ] then we can see that @xmath453 [ prop4.2 ] assume ( [ eq2.11 ] ) , ( [ eq2.12 ] ) , ( [ eq2.20 ] ) and ( [ eq2.22 ] ) .
then @xmath454 the proof of theorem [ theo2.2 ] follows directly from this proposition since @xmath455 because of proposition [ prop2.1 ] . for the proof of the present proposition ,
we prepare some lemmas . for each @xmath456
, we take the controlled process @xmath457 defined by ( [ eq4.8 ] ) and control @xmath458 .
taking a sequence @xmath459 such that @xmath460,\ ] ] we have the following lemma . under the assumptions of proposition [ prop4.2 ] , for each @xmath461 we have @xmath462=0.\ ] ]
set @xmath463+\bigl(g+\lambda{\hat z}(t , x)\bigr)^*d\psi.\ ] ] then @xmath464 therefore , @xmath465.\ ] ] since @xmath466\geq{\bar\chi},\ ] ] we have @xmath467=0.\ ] ] noting that @xmath468 and @xmath469 , because of lemma [ lem4.1 ] , we obtain @xmath470\leq0,\ ] ] and our lemma has been proved .
under the assumptions of proposition [ prop4.2 ] , there exists a subsequence @xmath471 such that @xmath472=0.\ ] ] @xmath473 & & { } + 2\int_{t - t}^t{\hat y}^{(t)}_s\bigl\{g\bigl({\hat y}^{(t)}\bigr)+\lambda\bigl({\hat y}_s^{(t)}\bigr){\hat z}\bigl(s,{\hat y}_s^{(t)}\bigr)\bigr\}\,ds\\[-2pt ] & & { } + \int_{t - t}^t\operatorname { tr}\bigl[\lambda\lambda\bigl({\hat y}_s^{(t)}\bigr)\bigr]\,ds . \ ] ] therefore , @xmath474&=&e\bigl[\bigl|{\hat y}_{t - t}^{(t)}\bigr|^2\bigr]+2e\biggl[\int _ { t - t}^t{\hat y}^{(t)}_s\bigl\{g\bigl({\hat y}^{(t)}\bigr)+\lambda\bigl({\hat y}_s^{(t)}\bigr){\hat z}\bigl(s,{\hat y}_s^{(t)}\bigr)\bigr\}\,ds \\[-2pt ] & & \hphantom{e\bigl[\bigl|{\hat y}_{t - t}^{(t)}\bigr|^2\bigr]+2e\biggl[}\hspace*{91pt } { } + \int_{t - t}^t\operatorname{tr}\bigl[\lambda\lambda\bigl({\hat y}^{(t)}_s\bigr)\bigr]\,ds\biggr ] . \ ] ] by using the gradient estimates in theorem [ theo2.1 ] and ( [ eq2.20 ] ) we obtain @xmath475 for some positive constant @xmath220 and @xmath476 & \leq&e\bigl[\bigl|{\hat y}_{t - t}^{(t)}\bigr|^2\bigr]+ce\biggl[\int _ { t - t}^t\bigl|{\hat y}_s^{(t)}\bigr|^2\,ds\biggr ] \\[-2pt ] & \leq&e\bigl[\bigl|{\hat y}_{t - t}^{(t)}\bigr|^2\bigr]+c'e\biggl[\int_{t - t}^t\varphi\bigl({\hat y}_s^{(t)},{\hat z}\bigl(s,{\hat y}_s^{(t)}\bigr)\bigr)\,ds\biggr ] \ ] ] since ( [ eq2.22 ] ) is assumed and @xmath477 & = & \frac{1}{2}(\lambda^*d{\bar v}-\sigma^*{\hat\alpha})^ * n_{\gamma } ( \lambda^*d{\bar v}-\sigma^*{\hat\alpha } ) \\[-2pt ] & & { } + ( \lambda^*d{\bar v}-\sigma^*{\hat\alpha } ) ^*n_{\gamma}^{-1}\sigma^*{\hat\alpha}+\frac{1}{2}{\hat\alpha } \sigma\sigma^*{\hat\alpha } \\[-2pt ] & = & \frac{1}{2}(d{\bar v})^*\lambda n_{\gamma}^{-1}\lambda^*d{\bar v}-\frac{\gamma}{2(1-\gamma)}{\hat\alpha}\sigma\sigma^*{\hat \alpha}. \ ] ] by it s formula , @xmath478 and we have @xmath479= e[{\bar v}(t - t,{\hat y}_{t - t};t)].\ ] ] take a subsequence @xmath480 such that @xmath481=0\ ] ] and @xmath482=0.\ ] ] then we have @xmath483&\leq&\lim_{t_n'\to\infty}\frac { 1}{t_n'}e\bigl[\bigl|{\hat y}_{t_n'-t}^{(t_n')}\bigr|^2\bigr ] \\ & & { } + c'\lim_{t_n'\to\infty}\frac{1}{t_n'}e\biggl[\int_{t_n'-t}^{t_n'}\varphi \bigl({\hat y}_s^{(t_n')},{\hat z}\bigl(s,{\hat y}_s^{(t_n')}\bigr)\bigr)\,ds\biggr]=0 , \ ] ] and thus the lemma has been proved .
proof of proposition [ prop4.2 ] for each @xmath248 there exists @xmath484 such that @xmath485 < { \varepsilon}t_{{\varepsilon}},\qquad{\bar w}(x ) < { \varepsilon}t_{{\varepsilon } } , & \\ & \displaystyle \biggl|{\bar\chi}-\frac{1}{t_{{\varepsilon}}}e\biggl[\int_0^{t_{{\varepsilon}}}\varphi \bigl({\hat y}_s^{(t_{{\varepsilon}})},{\hat z}_s^{(t_{{\varepsilon}})}\bigr)\,ds\biggr]\biggr|<{\varepsilon}.&\end{aligned}\ ] ] set @xmath486 and consider @xmath487 defined by @xmath488 then , for @xmath489 , @xmath490 and @xmath491\,dt.\end{aligned}\ ] ] therefore , @xmath492\bigr\}\,ds \\ & & \qquad \leq e^{p t_{{\varepsilon}}}\bigl|y_{t_{{\varepsilon}}}^{({\varepsilon})}\bigr|^2 + 2\int _ { t_{{\varepsilon}}}^te^{ps}\bigl(y_s^{({\varepsilon})}\bigr)^*\lambda\bigl(y_s^{({\varepsilon})}\bigr)\,dw_s+\int_{t_{{\varepsilon}}}^te^{ps}\bigl\{-k_1\bigl|y_s^{({\varepsilon})}\bigr|^2+k_2\bigr\}\,ds \ ] ] for some positive constants @xmath433 . by using stopping time arguments , for @xmath489 , we have @xmath493&\leq & e[e^{pt_{{\varepsilon}}}|y_{t_{{\varepsilon}}}|^2]+e\biggl[\int_{t_{{\varepsilon}}}^te^{ps}k_2 \,ds\biggr ] \\ & = & e\bigl[e^{pt_{{\varepsilon}}}\bigl|y_t^{({\varepsilon})}\bigr|^2\bigr]+\frac { k_2}{p}(e^{pt}-e^{pt_{{\varepsilon } } } ) . \ ] ] thus , we see that @xmath494\leq e\bigl[\bigl|y_{t_{{\varepsilon}}}^{({\varepsilon})}\bigr|^2\bigr]+\frac{k_2}{p},\ ] ] from which we obtain @xmath495=0 $ ] .
hence , @xmath496 .
now , by applying it s formula , we have @xmath497\bigl(y_s^{({\varepsilon})}\bigr)\,ds \\ & \geq&\int_0^{t_{{\varepsilon}}}\bigl(d{\bar w}\bigl(y_s^{({\varepsilon})}\bigr)\bigr)^ * \lambda \bigl(y_s^{({\varepsilon})}\bigr)\,dw_s-\chi t_{{\varepsilon}}-\int_0^{t_{{\varepsilon}}}\varphi\bigl(y_s^{({\varepsilon})},z_s^{({\varepsilon})}\bigr)\,ds . \ ] ] therefore , @xmath498,\ ] ] from which we have @xmath499+\frac{1}{t_{{\varepsilon}}}e\bigl[{\bar w}\bigl(y_{t_{{\varepsilon}}}^{({\varepsilon})}\bigr)\bigr ] \\ & \leq&{\varepsilon}+{\bar\chi}+c{\varepsilon}\ ] ] for some positive constant @xmath219 .
therefore , @xmath500 for any @xmath248 , and we have @xmath501 .
this completes the proof of the proposition .
the following is a direct consequence of proposition [ prop4.1 ] .
[ cor4.1 ] under the assumptions of proposition [ prop4.2 ] , @xmath502 is a concave function on @xmath503 , and @xmath31 is a convex function .
indeed , @xmath504 is a concave function of @xmath29 , and the infimum of a family of concave functions @xmath502 is concave .
[ prop4.3 ] under the assumptions of proposition [ prop3.1 ] , @xmath156 is ergodic . @xmath505 as @xmath506 , and @xmath507 and @xmath219 .
moreover , @xmath508 , and the hasminskii conditions ( cf .
@xcite ) hold .
we are going to consider a poisson equation formally obtained by differentiating h - j - b equation ( [ eq3.1 ] ) of ergodic type with respect to @xmath29 .
namely , we consider @xmath509+g^*du-(\lambda^*d{\bar w}-\sigma^*{\hat\alpha})^*n_{\gamma } ^{-1}\lambda^*du \\ & & { } -\frac{1}{2(1-\gamma)^2}(\lambda^*d{\bar w}-\sigma ^*{\hat\alpha})^*\sigma^*(\sigma\sigma^*)^{-1}\sigma(\lambda ^*d{\bar w}-\sigma^*{\hat\alpha } ) . \
] ] since @xmath510 we can write @xmath511 note that @xmath156 is ergodic in light of proposition [ prop4.3 ] , and the pair @xmath512 of a function @xmath513 and a constant @xmath514 is the solution to ( [ eq5.1 ] ) . let us set @xmath515 and let @xmath221 be sufficiently large so that @xmath516 \\[-8pt ] \eqntext{\displaystyle x\in{\mathcal d}^c,}\end{aligned}\ ] ] for @xmath517 , which is possible because of assumption ( [ eq2.22 ] ) .
therefore , we see that @xmath156 and @xmath209 satisfy assumption ( [ eq9.3 ] ) in the and also @xmath518 for @xmath519 in the following we always take a solution @xmath209 to ( [ eq3.1 ] ) such that @xmath520 . thus
, according to proposition [ prop9.4 ] we can show the existence of the solution @xmath512 to ( [ eq5.1 ] ) .
[ cor5.1 ] equation ( [ eq5.1 ] ) has a solution @xmath521 such that @xmath522 and @xmath523 moreover , this solution @xmath513 is unique up to additive constants .
it can be clearly seen that @xmath524 and proposition [ prop9.4 ] applies .
[ lem6.1][lem2.6 ] under the assumptions of proposition [ prop4.2 ] , @xmath525 where @xmath220 and @xmath262 are positive constants independent of @xmath526 . inequality ( [ eq6.1 ] ) is a direct consequence of ( [ eq4.15 ] ) in lemma [ lem4.2 ] since @xmath527 is an ergodic diffusion process with the invariant measure @xmath528 . in the following
, we always work under the assumptions of theorem [ theo2.2 ] ( proposition [ prop4.2 ] ) .
[ lem6.2 ] let @xmath529 and @xmath530 be solutions to ( [ eq3.1 ] ) with @xmath29 and @xmath531 , respectively , such that @xmath532 .
then @xmath533 converges to @xmath534 in @xmath535 strongly and uniformly on each compact set .
we have @xmath536 and @xmath537 for each @xmath183 in light of ( [ eq3.11 ] ) and ( [ eq3.15 ] ) , where @xmath538 is a positive constant independent of @xmath539 , @xmath540 .
therefore it follows that @xmath541 is bounded in @xmath542 and converges to some @xmath543 in @xmath542 weakly for each @xmath183 and also uniformly on each compact set by taking a subsequence if necessary .
note that @xmath544 converges to @xmath140 because @xmath140 is convex on @xmath545 and thus continuous .
take a function @xmath546 such that @xmath547 , and @xmath548 . with @xmath549
, we test @xmath550+\beta_{\gamma+\delta}^*d{\bar w}^{(\gamma+\delta ) } \\ & & { } -\tfrac{1}{2}\bigl(d{\bar w}^{(\gamma+\delta)}\bigr)^*\lambda n_{\gamma+\delta}^{-1}\lambda^*d{\bar w}^{(\gamma+\delta ) } + u_{\gamma+\delta } \ ] ] and obtain @xmath551 where @xmath552 .
therefore , @xmath553 since all terms of the right - hand side converge to @xmath554 , we see that converges strongly to @xmath554 in @xmath555 and @xmath533 to @xmath543 strongly in @xmath556 .
thus , we obtain our present lemma because @xmath557 satisfies ( [ eq3.1 ] ) , and the solution is unique up to additive constants with respect to @xmath209 . [ lem6.8][lem6.3 ] let @xmath558 be a solution to @xmath559 where @xmath560 then , as @xmath561 , @xmath562 converges to @xmath563 and @xmath564 converges to @xmath565 in @xmath535 strongly and uniformly on each compact set , where @xmath566 is a solution to ( [ eq5.1 ] ) . note that @xmath567 , and that @xmath568 almost everywhere by taking a subsequence , if necessary , since @xmath569 converges strongly in @xmath535 to @xmath570 by lemma [ lem6.2 ] .
moreover , we note that @xmath571 is tight because of lemma [ lem6.1 ]
. therefore , it converges weakly to some probability measure @xmath572 by taking a subsequence if necessary .
the limit can be identified with @xmath573 , and @xmath574 is the only function satisfying ( [ eq9.26 ] ) for @xmath575 and @xmath576 .
thus @xmath577 converges to @xmath578 weakly .
therefore , @xmath579 converges to @xmath563 . on the other hand , since @xmath564 is a solution to ( [ eq6.2 ] ) it satisfies @xmath580 and we have @xmath581 .
therefore , we see that @xmath564 is locally bounded by a constant independent of @xmath539 . then , by testing ( [ eq6.2 ] ) with @xmath582 , we can see that @xmath583 is bounded for each @xmath183 .
therefore , by taking a subsequence if necessary , @xmath564 converges in @xmath535 weakly to a function @xmath584 , which turns out to be the solution @xmath585 to ( [ eq5.1 ] ) . by similar arguments as in the proof of lemma [ lem6.2 ] ,
we can see it converges in @xmath535 strongly .
furthermore , by theorem 9.11 in @xcite , we see that @xmath586 for each @xmath587 , where @xmath118 is a constant depending on @xmath588 and the @xmath589 norms of the coefficients of @xmath590 .
thus , by the sobolev imbedding theorem , @xmath591 is equicontinuous , and thus @xmath564 converges uniformly to @xmath585 on each compact set . [ lem6.4 ] let @xmath529 and @xmath592 be solutions to ( [ eq3.1 ] ) with @xmath29 and @xmath531 , respectively . set @xmath593 and @xmath594 .
then @xmath595 and @xmath596 here , @xmath566 is the solution to ( [ eq5.1 ] ) .
here we abbreviate @xmath585 to @xmath513 . from ( [ eq3.1 ] )
it follows that @xmath597 & & \qquad = \tfrac{1}{2}\operatorname{tr}\bigl[\lambda \lambda^*d^2\bigl({\bar w}^{(\gamma+\delta)}-{\bar w}^{(\gamma)}\bigr)\bigr]+\beta _ { \gamma+\delta}^*d{\bar w}^{(\gamma+\delta ) } \\[-2pt ] & & \qquad \quad { } -\beta_{\gamma}^*d{\bar w}^{(\gamma)}-\tfrac { 1}{2}\bigl(d{\bar w}^{(\gamma+\delta)}\bigr)^*\lambda n_{\gamma+\delta } ^{-1}\lambda^*d{\bar w}^{(\gamma+\delta ) } \\[-2pt ] & & \qquad \quad { } + \tfrac{1}{2}\bigl(d{\bar w}^{(\gamma)}\bigr)^*\lambda n_{\gamma } ^{-1}\lambda^*d{\bar w}^{(\gamma)}+u_{\gamma+\delta}-u_{\gamma } \\[-2pt ] & & \qquad = \tfrac{1}{2}\operatorname{tr}\bigl[\lambda\lambda^*d^2\bigl({\bar w}^{(\gamma + \delta)}-{\bar w}^{(\gamma)}\bigr)\bigr]+\beta_{\gamma}^*d\bigl({\bar w}^{(\gamma + \delta)}-{\bar w}^{(\gamma)}\bigr ) \\[-2pt ] & & \qquad \quad { } -\bigl(d{\bar w}^{(\gamma)}\bigr)^*\lambda n_{\gamma } ^{-1}\lambda^*d\bigl({\bar w}^{(\gamma+\delta)}-{\bar w}^{(\gamma)}\bigr ) -\tfrac{1}{2}\bigl(d{\bar w}^{(\gamma)}\bigr)^*\lambda n_{\gamma}^{-1}\lambda ^*d{\bar w}^{(\gamma ) } \\[-2pt ] & & \qquad \quad { } + \bigl(d{\bar w}^{(\gamma)}\bigr)^*\lambda n_{\gamma } ^{-1}\lambda^*d{\bar w}^{(\gamma+\delta)}-\tfrac{1}{2}\bigl(d{\bar w}^{(\gamma+\delta)}\bigr)^*\lambda n_{\gamma+\delta}^{-1}\lambda ^*d{\bar w}^{(\gamma+\delta ) } \\[-2pt ] & & \qquad \quad { } + ( \beta_{\gamma+\delta}-\beta_{\gamma})^*d{\bar w}^{(\gamma + \delta)}+u_{\gamma+\delta}-u_{\gamma } \\[-2pt ] & & \qquad = { \bar l}(\gamma)\bigl({\bar w}^{(\gamma+\delta)}-{\bar w}^{(\gamma ) } \bigr)\\[-2pt ] & & \qquad \quad { } -\tfrac{1}{2}d\bigl({\bar w}^{(\gamma+\delta)}-{\bar w}^{(\gamma ) } \bigr)^*\lambda n_{\gamma+\delta}^{-1}\lambda^ * d\bigl({\bar w}^{(\gamma + \delta)}-{\bar w}^{(\gamma)}\bigr ) \\[-2pt ] & & \qquad \quad { } + ( \beta_{\gamma+\delta}-\beta_{\gamma})^*d{\bar w}^{(\gamma+\delta)}+\tfrac{1}{2}\bigl(d{\bar w}^{(\gamma)}\bigr)^*\lambda ( n_{\gamma+\delta}^{-1}-n_{\gamma}^{-1})\lambda^*d{\bar w}^{(\gamma ) } \\[-2pt ] & & \qquad \quad { } -\bigl(d{\bar w}^{(\gamma)}\bigr)^*\lambda ( n_{\gamma+\delta } ^{-1}-n_{\gamma}^{-1})\lambda^*d{\bar w}^{(\gamma+\delta ) } + u_{\gamma+\delta}-u_{\gamma}. \ ] ] therefore we have @xmath598 where @xmath599 and @xmath600 note that @xmath601 is dominated by @xmath602 with a certain positive constant @xmath220 and that it converges almost everywhere to @xmath603 by taking a subsequence , if necessary , because @xmath604 therefore , @xmath605 moreover , for @xmath606 , @xmath607 let us consider the equation @xmath608 we shall see that @xmath609 as @xmath610 by specifying suitable ambiguity constants . for that , set @xmath611 then we have @xmath612 by considering constructing the solution to this equation according to the proof of proposition [ prop9.4 ] in the , we see that @xmath613 as @xmath610 .
to begin , let @xmath614 be the solution to ( [ eq9.21 ] ) for @xmath615 , @xmath616 and @xmath617 the solution to ( [ eq9.22 ] ) for @xmath618 , @xmath616 and @xmath619 .
the operator @xmath0 is defined as @xmath620 and the operator @xmath621 is defined in the same way as in ( [ eq9.11 ] ) by replacing @xmath622 with @xmath575 in ( [ eq9.4 ] ) and ( [ eq9.10 ] ) . then starting with @xmath623 , @xmath624 , we define the sequence @xmath625 , @xmath626 successively as the solution to ( [ eq9.10 ] ) with @xmath627 and @xmath628 , and as the solution to ( [ eq9.4 ] ) with @xmath629 and @xmath618 , respectively .
then we obtain @xmath630 and the estimate for @xmath631 , @xmath632 to estimate @xmath633 , we set @xmath634 to be the solution to @xmath635 and @xmath636
. then @xmath637 satisfies @xmath638 and we have @xmath639 for some constant @xmath640 . on the other hand ,
to estimate @xmath641 , we set @xmath642 we can assume that @xmath643 is sufficiently large so that @xmath644 for some @xmath645 . since @xmath646 @xmath647 satisfies @xmath648 noting that @xmath649 we have @xmath650 and thus @xmath651 moreover , @xmath652 uniformly on each compact set as @xmath610 .
therefore , @xmath653 uniformly on each compact set and we also obtain the estimates @xmath654 and @xmath655 let @xmath656 be the solution to ( [ eq9.28 ] ) for @xmath618 , @xmath616 and @xmath657 .
then @xmath658 uniformly as @xmath610 since @xmath659 is estimated as shown above .
let @xmath660 be the solution to ( [ eq9.22 ] ) for @xmath618 , @xmath616 and @xmath661 . then ,
as above , @xmath662 uniformly on each compact set as @xmath610 .
since @xmath663 in @xmath664 and @xmath665 in @xmath666 , we conclude that @xmath667 uniformly on each compact set . in a similar manner
, we have @xmath668 by setting @xmath669 we have @xmath670 since @xmath671 converges to @xmath578 weakly , and @xmath672 converges almost everywhere to @xmath673 by taking a subsequence if necessary , as above , we have @xmath674 \\[-8pt ] \eqntext{\mbox{as } |\delta @xmath675 we consider @xmath676 then , in the same manner as above , we see that @xmath677 , as @xmath561 , by specifying ambiguity constants .
since @xmath564 converges to @xmath513 , @xmath678 does the same . from ( [ eq6.4 ] ) , ( [ eq6.5 ] ) , ( [ eq6.8 ] ) and ( [ eq6.9 ] )
, it follows that @xmath679 the converse inequalities of ( [ eq6.5 ] ) and ( [ eq6.9 ] ) hold for @xmath680 , and we have @xmath681 hence , we see that @xmath682 as @xmath561 . from ( [ eq6.3 ] ) and ( [ eq6.6 ] ) , we have @xmath683 and through arguments similar to those above , we see that @xmath684 since @xmath685 for @xmath606 and @xmath686 as @xmath561 .
similarly , from ( [ eq6.7 ] ) and ( [ eq6.10 ] ) , we have @xmath687 and see that @xmath688 therefore , @xmath689 we likewise have @xmath690 and @xmath691 since @xmath692 for @xmath680 .
therefore , we obtain @xmath693 and lemma [ lem6.4 ] follows .
[ rem6.1 ] since @xmath694 is a solution to ( [ eq5.1 ] ) , it has a polynomial growth order .
more precisely , we have @xmath695 cf . corollary [ cor5.1 ] and ( [ eq3.11 ] ) .
furthermore , we can see that @xmath696 also has a polynomial growth order for each @xmath697 .
indeed , @xmath698 satisfies @xmath699 where @xmath700 therefore , if we set @xmath701 , and @xmath702 , then @xmath703 satisfies @xmath704 for each @xmath705 , @xmath706 .
we note that @xmath707 which can be seen in light of our assumptions since @xmath513 is a solution to ( [ eq5.1 ] ) , and @xmath209 is a solution to ( [ eq3.1 ] ) .
equation ( [ eq6.14 ] ) corresponds to ( 13.4 ) in @xcite , chapter 3 , section 13 .
therefore , by the same arguments as in that work , @xmath708\ ] ] is seen to hold , where @xmath709 is a cut - off function supported by @xmath710 , @xmath711 , and @xmath712 is a constant dominated by @xmath713 , @xmath422 , @xmath714 . from this inequality
we obtain inequality ( 5.12 ) for @xmath715 in @xcite , chapter 2 , section 5 .
hence , similarly to the proof of lemma 5.4 in @xcite , chapter 2 , we see that @xmath703 has a polynomial growth order .
we first state the following lemma . [ lem7.1 ] under the assumptions of theorem [ theo2.3 ] , @xmath716 note that @xmath717 for @xmath290 and that @xmath149 is nondecreasing
. therefore , @xmath718 exists . furthermore , @xmath719 and @xmath720 since @xmath721 . here
, @xmath722 is a constant defined by ( [ eq3.10 ] ) .
hence , we obtain ( [ eq7.1 ] ) .
we next give the proof of theorem [ theo2.4 ] .
for @xmath23 , we have @xmath723e^{-\gamma\kappa t}\biggr\ } \\ & = & \log e\biggl[\biggl(\frac{v_t(h)}{s_t^0}\biggr)^{\gamma}\biggr ] -\gamma\kappa t. \ ] ] therefore , @xmath724 -\gamma\kappa t\\ & \leq & v(0,x;t)-\gamma\kappa t,\end{aligned}\ ] ] from which we obtain @xmath725 for all @xmath23 .
hence , we have @xmath726 the converse inequality is more difficult to prove . take a constant @xmath15 and @xmath727 such that @xmath728
. then there exists @xmath729 such that @xmath730 we write @xmath729 as @xmath29 for simplicity in the following .
let us introduce a probability measure @xmath731 defined by @xmath732 where @xmath733 then @xmath734 is a martingale under the probability measure @xmath731 and @xmath735 \\[-8pt ] & = & \biggl\{\beta+\frac{\gamma}{1-\gamma}\lambda\sigma^*{\hat\alpha}+ \lambda n^{-1}_{\gamma}\lambda^*dw\biggr\}(x_t)\,dt+\lambda(x_t)\,d{\tilde w}_t . \nonumber \ ] ] note that @xmath736 has the same law as the diffusion process @xmath737 governed by stochastic differential equation ( [ eq4.12 ] ) .
we further note h - j - b equation of ergodic type ( [ eq2.19 ] ) can be written as @xmath738 by using @xmath156 defined by ( [ eq4.16 ] ) . on the other hand , ( [ eq5.1 ] )
is written as @xmath739 owing to lemma [ lem6.3 ] , where @xmath740 .
now we have @xmath741 and we set @xmath742 note that @xmath743 .
then it follows that @xmath744 taking lemma [ lem4.2 ] into account , we have @xmath745 & & \qquad \leq\frac{9}{{\varepsilon}^2 t^2}{\tilde e}\biggl[\int _ 0^tv_1(x_s)\,ds\biggr ] \\[-2pt ] & & \qquad \leq\frac{c}{{\varepsilon}^2 t } \ ] ] for some positive constant @xmath118 and @xmath746\leq e^{-{{{\varepsilon}t}/{3}}}.\ ] ] thus , by using the following lemma , we can see that @xmath747 for sufficiently large @xmath0 .
[ lem7.2 ] for sufficiently large @xmath0 we have @xmath748 by it s formula , @xmath749 & & \qquad = \int_0^t\{{\bar l}(\gamma)w_{\gamma}(x_s)\}\,ds+\int
_ 0^t(\nabla w_{\gamma}(x_s))^*\lambda(x_s)\,d{\tilde w}_s \\[-2pt ] & & \qquad = -\int_0^tv_1(x_s)\,ds+\chi'(\gamma)t+\int_0^t(\nabla w_{\gamma}(x_s))^*\lambda(x_s)\,d{\tilde w}_s . \ ] ] therefore , @xmath750 & & { } + \frac{1}{t}\int_0^t(\nabla w_{\gamma } ( x_s))^*\lambda(x_s)\,d{\tilde w}_s.\end{aligned}\ ] ] thus , @xmath751 & & \qquad = { \tilde p}\biggl(\frac{1}{t}\log v+\frac{1}{t}\{w_{\gamma } ( x)-w_{\gamma}(x_t)\}+ \frac{1}{t}\int_0^t(\nabla w_{\gamma}(x_s))^*\,d{\tilde w}_s>\frac { { \varepsilon}}{3}\biggr ) \\[-2pt ] & & \qquad \leq{\tilde p}\biggl(\frac{1}{t}\log v>\frac{{\varepsilon}}{9}\biggr)+{\tilde p}\biggl(\frac{1}{t}\{w_{\gamma}(x)-w_{\gamma}(x_t)\}>\frac { { \varepsilon}}{9}\biggr ) \\[-2pt ] & & \qquad \quad { } + { \tilde p}\biggl(\frac{1}{t}\int_0^t(\nabla w_{\gamma}(x_s))^*\,d{\tilde w}_s>\frac{{\varepsilon}}{9}\biggr ) \\[-2pt ] & & \qquad \leq\frac{81}{{\varepsilon}^2t^2}e[|w_{\gamma}(x)-w_{\gamma}(x_t)|^2 ] + \frac{81}{{\varepsilon}^2t^2}e\biggl[\int_0^t(d w_{\gamma})^*\lambda\lambda ^*dw_{\gamma}(x_s)\,ds\biggr ] . \ ] ] hence , by taking @xmath0 and @xmath237 to be sufficiently large , we obtain our present lemma because of lemma [ lem4.2 ] ; cf . remark [ rem6.1 ] .
let us complete the proof of theorem [ theo2.4 ] for @xmath752 .
set @xmath753 and @xmath754 a_2&=&\bigl\ { -\tfrac{1}{2}\langle m^{\gamma}\rangle_t\geq\bigl(\chi(\gamma ) -\gamma\chi'(\gamma)-{\varepsilon}\bigr)t\bigr\ } , \\[-2pt ] a_3&=&\biggl\ { \frac{1}{t}\bigl(\log v_t(h)-\log s_t^0\bigr)\leq\kappa\biggr\}. \ ] ] then @xmath755 & & \qquad = { \tilde e}\biggl[e^{-{\tilde m}^{\gamma}_t-{({1}/{2})}\langle m^{\gamma}\rangle_t};\frac { 1}{t}\bigl(\log v_t(h)-s_t^0\bigr)\leq\kappa\biggr ] \\[-2pt ] & & \qquad \geq{\tilde e}\bigl[e^{-{\tilde m}^{\gamma}_t-{({1}/{2})}\langle m^{\gamma}\rangle_t } ; a_1\cap a_2\cap a_3\bigr ] \\[-2pt ] & & \qquad \geq e^{(\chi(\gamma)-\gamma\chi'(\gamma)-2{\varepsilon})t}{\tilde p}(a_1\cap a_2\cap a_3 ) \\[-2pt ] & & \qquad \geq e^{(\chi(\gamma)-\gamma\chi'(\gamma)-2{\varepsilon})t}\ { 1-{\tilde p}(a_1^c)-{\tilde p}(a_2^c)-{\tilde p}(a_3^c)\}. \ ] ] we have seen that @xmath756 holds for sufficiently large @xmath0 in ( [ eq7.6 ] ) , and it is straightforward to see that likewise @xmath757 for sufficiently large @xmath0 .
therefore , taking the following lemma into account as well , we have @xmath758 for sufficiently large @xmath0 , and @xmath759 & = & \chi(\gamma)-\gamma(\kappa-{\varepsilon})-2{\varepsilon}\\[-2pt ] & \geq & \inf_{\gamma<0}\{\chi(\gamma)-\gamma(\kappa-{\varepsilon})\ } -2{\varepsilon}\ ] ] for each @xmath248 . since @xmath140 is smooth and convex , @xmath760 , @xmath20 , is strictly concave , and thus continuous .
hence , @xmath761 under the assumptions of theorem [ theo2.4 ] , @xmath762 holds for sufficiently large @xmath0 .
first note that @xmath763 and @xmath764 then , from ( [ eq7.4 ] ) and ( [ eq7.5 ] ) , we have @xmath765 set @xmath766 then @xmath767 by it s formula , we have @xmath768 thus , @xmath769 hence , we obtain the present lemma in the same way as lemma [ lem7.2 ] . for @xmath163 , we shall prove theorem [ theo2.4 ] . by convexity
, we have @xmath770 that is , @xmath771 @xmath149 is monotonically nondecreasing , and @xmath772 as @xmath773 .
therefore , we see that @xmath774 hence , @xmath775 on the other hand , by taking @xmath776 , we have @xmath777 and @xmath778 for sufficiently large @xmath0 .
thus , @xmath779 .
for a given constant @xmath752 , take @xmath159 such that @xmath780 , namely , @xmath781 then , since @xmath782-\gamma ( \kappa)\kappa t,\ ] ] we have @xmath783-\gamma(\kappa)\kappa.\ ] ] therefore , if we prove that @xmath784&=&\lim_{t\to\infty}\frac { 1}{t}\log e\biggl[\biggl(\frac{v_t(h^{(\gamma(\kappa))})}{s_t^0}\biggr)^{\gamma ( \kappa)}\biggr]\nonumber \\[-8pt ] \\[-8pt ] \quad & = & \chi(\gamma(\kappa ) ) , \nonumber\end{aligned}\ ] ] then we complete the proof of the present theorem because @xmath785-\gamma(\kappa)\kappa\ ] ] and @xmath760 by theorem [ theo2.4 ] .
( [ eq8.1 ] ) is proved in the following proposition . under the assumptions of theorem [ theo2.5 ] , ( [ eq8.1 ] )
let @xmath786 be a solution to ( [ eq2.19 ] ) for @xmath787 and @xmath788 , where @xmath123 is the solution to ( [ eq2.25 ] ) .
noting that @xmath789 \\[-8pt ] & = & \frac{1}{2(1-\gamma)}{\hat\alpha}^*\sigma\sigma^*{\hat\alpha } -\frac{1}{2(1-\gamma)}(dw)^*\lambda\sigma^*(\sigma\sigma ^*)^{-1}\sigma\lambda^*dw , \nonumber \ ] ] we have @xmath790+(\beta+\gamma\lambda\sigma^*{\bar h})^*dw\biggr\}(x_s)\,ds \\ & & \qquad \quad { } + \int_0^t(dw)^*\lambda(x_s)\,dw_s^{{\bar h } } \\ & & \qquad = \int_0^t\biggl\{\frac{1}{2}\operatorname{tr}[\lambda\lambda^*d^2w]+\beta _ { \gamma}^*dw\\ & & \qquad\hphantom{=\int_0^t\biggl\ { } { } + \frac{\gamma}{1-\gamma}(dw)^*\lambda\sigma^*(\sigma \sigma^*)^{-1}\sigma\lambda^*dw\biggr\}(x_s)\,ds \\ & & \qquad \quad { } + \int_0^t(dw)^*\lambda(x_s)\,dw_s^{{\bar h } } \\ & & \qquad = \int_0^t\biggl\{\chi+u_{\gamma}+\frac{\gamma}{2(1-\gamma ) } ( dw)^*\lambda\sigma^*(\sigma\sigma^*)^{-1}\sigma\lambda ^*dw\\ & & \hspace*{179.5pt}{}-\frac{1}{2}(dw)^*\lambda\lambda^*dw\biggr\}(x_s)\,ds \\ & & \qquad \quad { } + \int_0^t(dw)^*\lambda(x_s)\,dw_s^{{\bar h } } \\ & & \qquad = \int_0^t\biggl\{\chi-\gamma\eta\bigl(x_s,{\bar h}_s^{(\gamma)}\bigr)-\frac { 1}{2}(dw)^*\lambda\lambda^*dw(x_s)\biggr\}\,ds\\ & & \qquad \quad { } + \int_0^t(dw)^*\lambda ( x_s)\,dw_s^{{\bar h}}. \ ] ] thus , @xmath791 \\ & & \qquad = e^{{\bar h}}\bigl[e^{\chi t+w(x)-w(x_t)+\int_0^t(dw)^*\lambda ( x_s)\,dw_s^{\bar h}-{({1}/{2})}\int_0^t(dw)^*\lambda\lambda^*dw(x_s)\,ds}\bigr ] . \ ] ] let us introduce a new measure @xmath792 defined by @xmath793 then @xmath794 is a brownian motion process under @xmath792 and @xmath795 therefore , @xmath796 & & \qquad = -\int_0^te^{-w(x_s)}(dw)^*\lambda ( x_s)\,d{\check w}_s \\[-0.5pt ] & & \qquad \quad { } + \int_0^te^{-w(x_s)}\biggl\{-\frac{1}{2}\operatorname{tr}[\lambda\lambda ^*d^2w]-(\beta+\gamma\lambda\sigma^*{\bar h})^*dw\\[-0.5pt ] & & \hspace*{183pt}{}-\frac { 1}{2}(dw)^*\lambda\lambda^*dw\biggr\}(x_s)\,ds \\[-0.5pt ] & & \qquad = -\int_0^te^{-w(x_s)}(dw)^*\lambda(x_s)\,d{\check w}_s \\[-0.5pt ] & & \qquad \quad { } -\int_0^te^{-w(x_s)}\biggl\{\frac{1}{2}\operatorname{tr}[\lambda\lambda ^*d^2w]+\beta_{\gamma}^*dw\\[-0.5pt ] & & \hphantom{\qquad \quad { } -\int_0^te^{-w(x_s)}\biggl\{}{}+\frac{\gamma}{1-\gamma}(dw)^*\lambda \sigma^*(\sigma\sigma^*)^{-1}\sigma\lambda^*dw \\[-0.5pt ] & & \hspace*{149pt } \qquad \quad { } + \frac{1}{2}(dw)^*\lambda\lambda^*dw\biggr\}(x_s)\,ds \\[-0.5pt ] & & \qquad = -\int_0^te^{-w(x_s)}(dw)^*\lambda(x_s)\,d{\check w}_s-\int _ 0^te^{-w(x_s)}\bigl\{\chi-\gamma\eta\bigl(x_s,{\bar h}_s^{(\gamma)}\bigr)\bigr\}\,ds . \ ] ] then , by the arguments using the stopping time , we have @xmath797\\[-0.5pt ] & & \qquad = e^{\chi t+w(x)}{\check e}\bigl[e^{-w(x_t)}\bigr ] \\[-0.5pt ] & & \qquad = e^{\chi t+w(x)}{\check e}\biggl[e^{-w(x)}+\int_0^te^{-w(x_s)}\bigl\{\gamma \eta\bigl(x_s,{\bar h}_s^{(\gamma)}\bigr)-\chi\bigr\}\,ds\biggr ] . \ ] ] hence , we obtain @xmath798\leq\chi(\gamma)\ ] ] by taking into account ( [ eq2.26 ] ) and ( [ eq8.2 ] ) .
the converse inequality holds since @xmath799 .
let @xmath622 be an elliptic operator defined by @xmath800 \\[-8pt]&\hspace*{2.3pt}=&\frac { 1}{2}\sum_{i , j}d_i(a^{ij}(x)d_ju)+\sum_i{\tilde b}^i(x)d_iu , \nonumber\end{aligned}\ ] ] where @xmath801 and @xmath802 are lipschitz continuous functions such that @xmath803 and @xmath804 .
we assume that there exists a positive function @xmath805 such that @xmath806 set @xmath807 , @xmath808 and @xmath809 then we consider the following exterior dirichlet problem for a given bounded continuous function @xmath16 on @xmath810 : @xmath811 [ prop9.1 ] exterior dirichlet problem ( [ eq9.4 ] ) has a unique bounded solution @xmath812 , @xmath813 .
we first show uniqueness .
note that @xmath814 and set @xmath815 .
then @xmath816 therefore , @xmath107 satisfies @xmath817 let @xmath818 and @xmath819 be solutions to ( [ eq9.5 ] ) .
then @xmath820 satisfies @xmath821 to prove uniqueness , it is sufficient to show that the solution @xmath822 to ( [ eq9.6 ] ) is trivial . for each @xmath727
, there exists @xmath823 such that @xmath824 .
take @xmath825 .
then we see that @xmath826 since @xmath827 and @xmath828 in @xmath666 .
thus , we see that @xmath829 because @xmath248 is arbitrary .
let us show the existence of the solution to ( [ eq9.4 ] ) .
we can assume @xmath830 .
consider the following dirichlet problem for @xmath222 : @xmath831 then we have @xmath832 it is clear that @xmath833 and @xmath238 by the maximum principle .
therefore , there exists @xmath834 and @xmath835 when taking @xmath836 we see that @xmath837 thus , @xmath838 converges to @xmath839 weakly in @xmath840 .
regularity theorems show that @xmath841 .
let us take a bounded domain @xmath664 such that @xmath842 and a bounded borel function @xmath843 on @xmath844 .
we consider a dirichlet problem @xmath845 which admits a solution @xmath846 .
for this solution , we consider exterior dirichlet problem ( [ eq9.4 ] ) with @xmath847 .
then we introduce an operator @xmath848 defined by @xmath849 where @xmath850 is the solution to ( [ eq9.4 ] ) with @xmath847 . in a similar manner to lemma 5.1 in @xcite , chapter ii
, we have @xmath851 where @xmath852 moreover , we have the following proposition ; cf .
theorem 4.1 , chapter ii in @xcite .
operator @xmath621 defined above satisfies the following properties .
@xmath853 and for some @xmath421 , @xmath854 furthermore , there exists a probability measure @xmath855 on @xmath856 such that @xmath857 \\[-8pt ] \eqntext{\displaystyle \rho=\log\frac{1}{1-\delta } , k=\frac{2}{1-\delta},}\end{aligned}\ ] ] and @xmath858 for any bounded borel function @xmath843 .
consider an exterior dirichlet problem for a given function @xmath271 : @xmath859 then we have the following proposition .
for a given function @xmath271 , there exists a unique solution @xmath860 , @xmath813 , to ( [ eq9.16 ] ) such that @xmath861 assume that @xmath862 , @xmath271 .
for @xmath222 we consider a dirichlet problem on @xmath863 : @xmath864 there exists a unique solution @xmath865 . set @xmath866
then we have @xmath867 to see that , set @xmath868 .
then @xmath869 therefore , @xmath870 since @xmath871 is subharmonic in @xmath872 and @xmath873 on @xmath874 .
hence , we have @xmath875 . on the other hand , @xmath876 hence , @xmath877 is superharmonic and @xmath878 on @xmath879 .
thus , @xmath880 and ( [ eq9.18 ] ) holds .
if @xmath881 and @xmath882 , then , through the same arguments for @xmath883 , we obtain @xmath884 where @xmath885 is the corresponding solution to ( [ eq9.17 ] ) .
therefore , for general @xmath886 , we have @xmath887 let @xmath888 be a solution to ( [ eq9.17 ] ) for @xmath889 .
then , @xmath888 is nondecreasing with respect to @xmath237 because of the maximum principle .
indeed , for @xmath238 , we have @xmath890 since @xmath888 is dominated by @xmath891 , there exists @xmath892 such that @xmath893 set @xmath894 then we have @xmath895 for @xmath896 injection @xmath897 , is compact .
therefore , @xmath898 weakly in @xmath840 for each @xmath899 and @xmath892 is a weak solution to @xmath900 by the regularity theorem @xmath901 , @xmath902 .
similarly , we have @xmath903 , which is a solution to @xmath904 now let us prove uniqueness . for @xmath540
, we assume that @xmath905 is a solution to ( [ eq9.16 ] ) such that @xmath906 then @xmath907 satisfies @xmath908 we shall prove that @xmath513 satisfying ( [ eq9.19 ] ) is trivial , @xmath909 . for this purpose ,
we set @xmath910 where @xmath911 is the constant that appears in ( [ eq9.3 ] ) . since @xmath912 we have @xmath913 note that @xmath914 for @xmath915 under assumption ( [ eq9.3 ] ) .
moreover , @xmath916 hence , from the maximum principle , we see that @xmath227 as in the proof of proposition [ prop9.1 ] .
let @xmath917 be a function on @xmath918 such that @xmath917 is bounded in @xmath919 and @xmath920 , and @xmath664 a bounded domain such that @xmath921 .
we consider @xmath922 and @xmath923 then we set @xmath924 and @xmath925 we further consider @xmath926 then , as in the proof of theorem 5.3 , in @xcite , chapter ii , we obtain the following proposition . here , we only give the proof of the existence of the solution for use in section [ sec6 ] .
[ prop9.4 ] equation ( [ eq9.24 ] ) has a solution unique up to additive constants if and only if @xmath927 .
moreover , @xmath928 for @xmath929 , @xmath930 and @xmath931 in distribution sense @xmath932 such that @xmath933 and @xmath934 .
furthermore , @xmath935 is the only function in @xmath936 satisfying ( [ eq9.26 ] ) and @xmath937 proof of existence let @xmath938 and @xmath939 , where @xmath940 ( resp . , @xmath839 )
is the solution to ( [ eq9.21 ] ) [ resp . , ( [ eq9.22 ] ) ] . for each @xmath941 define @xmath942 and @xmath943 as follows
let @xmath942 be the solution to ( [ eq9.10 ] ) for @xmath944 , and @xmath943 the solution to ( [ eq9.4 ] ) for @xmath945 .
then @xmath946 since @xmath947 , we have @xmath948 by ( [ eq9.14 ] ) .
set @xmath949 .
then @xmath950 therefore , we see that there exists @xmath951 such that @xmath952 , .
moreover , we have @xmath953 note that @xmath954 is the solution to @xmath955 and @xmath956 the solution to ( [ eq9.22 ] ) with @xmath957 . noting that @xmath958 we see that @xmath954 converges in @xmath959 and weakly in @xmath840 since its @xmath960 norm is bounded ; cf .
theorem 9.11 in @xcite . by the regularity theorems , the limit @xmath961 and satisfies @xmath962 on the other hand , @xmath963 is the solution to ( [ eq9.4 ] ) with @xmath964 and @xmath965 thus , we see that @xmath966 and @xmath956 converges in @xmath967 and weakly in @xmath968 .
the limit @xmath969 and satisfies ( [ eq9.22 ] ) with @xmath970 . setting @xmath971 in @xmath664 and @xmath972 in @xmath666
, we have a solution to ( [ eq9.24 ] ) .
the author thanks the anonymous referee for careful reading the manuscript and giving valuable suggestions .
the author also thanks dr .
y. watanabe for fruitful discussions . | we consider minimizing the probability of falling below a target growth rate of the wealth process up to a time horizon @xmath0 in an incomplete market model , and then study the asymptotic behavior of minimizing probability as @xmath1 .
this problem can be closely related to an ergodic risk - sensitive stochastic control problem in the risk - averse case . indeed , in our main theorem
, we relate the former problem concerning the asymptotics for risk minimization to the latter as its dual . as a result
, we obtain an expression of the limit value of the probability as the legendre transform of the value of the control problem , which is characterized as the solution to an h - j - b equation of ergodic type , in the case of a markovian incomplete market model . . |
the study of interfaces is often used to investigate the stability of ordered coexisting phases .
low temperature ordered phases are spatially homogeneous , however , interfaces can be induced by forcing heterogeneous boundary conditions @xcite .
for example , in the case of phases characterized by different values of an order parameter , one can fix the order parameter at the boundaries along one selected direction to two different values . for stable phases
the induced interface costs an amount of free - energy proportional to a power of the interface size . at the lower critical dimension this power vanishes and long - range order becomes unstable . in spin glasses , due to the lack of a physical order parameter allowing to distinguish the different phases , this definition of an interface does not seem of practical use .
different proposals have been put forward to define interfaces in spin glasses .
a common procedure @xcite , involves the comparison of free - energies between systems with periodic and antiperiodic boundary conditions , respectively .
this leads to a definition of an interface exponent , from the relation : @xmath0 which plays a fundamental rle in the droplet picture of spin glasses @xcite .
measures of @xmath1 have been used to perform a high - precision estimate of the spin - glass lower critical dimension @xmath2 in ref .
@xcite ( defined as the value of @xmath3 for which the interface exponent @xmath1 vanishes ) .
an alternative definition @xcite uses the free - energy cost that is needed to impose a spatial heterogeneity in the order parameter , which in spin glasses is the overlap , i.e. a measure of the similarity between two equilibrated configurations of the same system @xcite . one
needs to consider two identical copies of the system constrained in such a way to have their mutual overlaps on the boundaries fixed to some preassigned values .
this definition is well suited to test an important property emerging in the replica symmetry breaking ( rsb ) of the spin - glass phase : while the probability distribution @xmath4 for the overlap is broad
although having noticeable tails for finite systems reduces to two delta functions in the thermodynamic limit . ] , reflecting the presence of many different and almost degenerate equilibrium phases , one has that for any couple of equilibrated configurations their mutual overlap is spatially homogeneous on a large scale .
two configurations with a given value of the overlap in a given macroscopic portion of space will have the same overlap everywhere .
starting from this observation a pair of identical systems subject to `` twisted overlap '' boundary conditions was considered in @xcite .
two different values @xmath5 and @xmath6 of the overlap were imposed on the two boundaries along a selected direction , chosen among the ones with non - zero probability density in the @xmath4 .
it was then argued , that if rsb is present , the boundaries induce a smooth overlap profile in space that interpolates between the two values . above the lower critical dimension @xmath7 for spin - glass order this has a diverging free - energy cost , and one can expect the following scaling with @xmath8 and @xmath9 : @xmath10 where the exponent @xmath11 is positive for @xmath12 . a mean - field calculation of @xmath13 and the interface exponent @xmath14 gave the non - trivial values of @xmath15 and @xmath16 .
remarkably , while measuring a quite different property , the resulting value of the lower critical dimension is in very good agreement with the one estimated by the periodic - antiperiodic boundary condition method @xcite .
unfortunately , despite its potentially informative content on the nature of the spin - glass phase , the definition of @xcite was never used in numerical studies of the spin - glass phase .
this is probably due to the difficulty of imposing values of the overlap at the boundaries . in this paper
we start from the general observation , in fact not at all specific to spin glasses , that it is possible to induce interfaces with an alternative procedure
. one can divide the physical system into two contiguous halves and impose fixed values to the bulk order parameters in each of the two halves .
this procedure , theoretically equivalent to fixing heterogeneous boundary conditions , is better suited for numerical investigations , since the resulting free - energy cost can be related to the probability of a spontaneous fluctuation in an unconstrained system . in the case of spin glasses we have to consider two real replicas and impose values @xmath5 and
@xmath6 of their mutual overlaps in two contiguous half spaces .
this way of imposing heterogeneity in the system induces an overlap profile in space equal to the one induced by fixing the overlap on the boundaries to suitable values @xmath17 and @xmath18 .
one then has a free - energy cost @xmath19 of the form ( [ uno ] ) with the same exponents @xmath14 and @xmath13 .
the free - energy cost for imposing the different overlaps is related by boltzmann s relation to the probability that a spontaneous fluctuation of a couple of unconstrained systems produce the values @xmath5 and @xmath6 of the overlaps in the two half spaces : @xmath20 , which is a large deviation formula .
this quantity has the advantage that it is easily accessible in numerical simulations and therefore we base our analysis on it .
in this paper we focus our attention on overlap interfaces in dyson - like hierarchical spin - glass models .
hierarchical models without disorder @xcite have played an important rle in the theoretical and mathematical understanding of critical points @xcite . in the ferromagnetic dyson model
one can write exact renormalization group equations that involve the iteration of a function of a single variable ( for a review see @xcite ) .
hierarchical spin - glass models of the dyson kind have not to our knowledge been considered in the literature and are in our opinion very attractive as they could allow for an analytical study of non - mean - field disordered critical points and nontrivial low temperature phases .
these models provide a hierarchical version of spin glasses with power - law interactions introduced in @xcite .
one - dimensional models with power - law interactions have recently received a great deal of attention as test grounds for theoretical ideas about finite - range spin - glass phases @xcite . here , following dyson , we propose to use the tree topology as a further simplification . we choose to start the study of these models focusing on interface properties , for which we can get particularly simple theoretical predictions in the rsb framework .
we study the system through the replica method , deriving a recursion equation relating the replica partition functions at the different levels of the hierarchy .
this relation , which in principle codes for all the thermodynamic properties of the system is analyzed in a self - consistent way to obtain the interface free - energy .
we test our prediction in numerical simulations , finding good agreement .
the organization of the paper is the following : in section [ sec : general ] we discuss the definition of the interfaces and how in principle interfaces can be evaluated and some expectations based on rsb theory .
section [ sec : hierarchical ] is devoted to the definition of the models we use . in section [ sec : interfaces ] we discuss possible scenarios and expectations and in section [ sec : derivation ] we sketch the theoretical derivation of the interface free - energy cost in hierarchical models . section [ sec : numerics ] discusses the numerical simulations .
finally , section [ sec : summary ] concludes the paper . in the appendix
we discuss the numerical characterization of the critical point .
in this section we define the probability distribution of overlaps in contiguous half spaces and discuss how it can be computed in principle in the large deviation regime . for definiteness
we consider a spin - glass system with ising spins @xmath21 on a set of indices @xmath22 with @xmath23 elements , that we divide in two `` half spaces '' @xmath24 and @xmath25 with @xmath26 elements each . for two spin configurations @xmath27 and @xmath28 one can define the partial overlaps @xmath29 in each of the half spaces @xmath30 , @xmath31 as @xmath32 .
the joint probability distribution function ( pdf ) of the two overlaps , for fixed value of the system size @xmath8 and quenched disorder @xmath33 , is given by @xmath34 where @xmath35 denotes the partition function . from this relation one
gets the usual pdf of the total overlap by simple integration .
one can consider the average over the disorder of ( [ p12 ] ) , however , here we will concentrate on the large deviation regime , where one can expect @xmath36 with a positive @xmath37 .
the exponent @xmath38 represents an interface free - energy cost to maintain the constrained values . in this regime ,
the large deviation functional can be expected to be self - averaging and one needs to compute the average free - energy @xmath39 the rsb implies that if one chooses @xmath40 in the domain where @xmath4 is non - zero , then @xmath41 .
the property of homogeneity of the overlap should then translate in a form for @xmath19 of the kind @xmath42 . from the complete knowledge of the joint distribution ( [ joint ] ) one can in principle extract the marginal distribution of the difference @xmath43 @xmath44 in the large deviation regime , where this analysis is supposed to be valid , we have @xmath45 for @xmath46 and hence the integral should be dominated by the value of @xmath47 that maximizes the function @xmath48 . as a consequence @xmath49 , the pdf of @xmath50 , should behave as @xmath51 in the tails . in order to understand the behavior for `` small '' values of @xmath50 we can suppose a smooth cross - over between the small and the large fluctuation regimes . in this case the form ( [ log ] ) suggests a finite probability for @xmath52 , i.e. , a scaling form @xmath53 and the marginal probability of the difference @xmath43 , @xmath54 .
this quantity @xmath49 is particularly easily accessible in numerical simulations and contains the the information on the ratio between the interface exponents .
the comparison of this method with the study of the overlap correlation function @xcite , as well as the comparison of our analytic results in section 5 with pertubative calculations @xcite will be not attempted in this paper .
here we introduce hierarchical spin - glass models on dyson lattices .
we will consider two families of models , a first one that is better suited for analytic studies and a second one that is more adapted to numerical simulations . in both families
the spins are associated to the leaves of a binary tree , see fig .
( [ figuno ] ) .
the distance between two spins that , rising up in the hierarchy , meet after @xmath55 branches is naturally defined as @xmath56 .
the first family of models we would like to define , is the natural spin - glass generalization of the ferromagnetic dyson model ( @xcite ) .
the hamiltonian can be constructed iteratively , connecting the two non - interacting systems of @xmath57 spins @xmath58 to form a composite system of @xmath59 spins in the following way : @xmath60 & = h_{k}^{j_1}[s_1, .... ,s_{2^{k}}]+h_{k}^{j_2}[s_{2^k+1}, .... ,s_{2^{k+1 } } ] - \nonumber\\ & - \frac{1}{2^{(k+1)\sigma}}\sum_{i < j}^{1,2^{k+1}}j_{ij}s_i s_j \label{i}\end{aligned}\ ] ] having defined the hamiltonian for a system at the first level of the hierarchy @xmath61 the couplings @xmath62 are independent and identically distributed gaussian random variables with zero mean and unit variance .
the family is parametrized by the value of @xmath63 that tunes the decrease of the strength of the interaction with the distance .
the interaction strenght decrease as a power of the distance , the model is therefore a hierarchical counterpart of the one - dimensional spin glass with power - law interactions @xcite which has received attention recently @xcite . as in this case
, the model is defined for @xmath64 .
exactly at the value @xmath65 it reduces to the infinite range model @xcite . notice that the sum of the squares of the interaction terms that couple the two subsystems @xmath66 scales as @xmath67 .
this is on the order of the volume for the mean - field value @xmath68 and smaller than the volume for the nontrivial regime @xmath69 .
the second family we consider is a dilute version of model ( [ i ] ) .
we consider a hamiltonian for @xmath57 spins with a fixed number of terms @xmath70
@xmath71=-\sum_{\mu=1}^p j^\mu s_{i^\mu } s_{j^\mu } , \label{ii}\end{aligned}\ ] ] where the @xmath72 are taken as @xmath73 variables with equal probability , and we choose the interacting couples of sites @xmath74 independently term by term and in a way that , if the distance between @xmath75 and @xmath76 on the binary tree with @xmath57 branches is @xmath77 , we put the coupling with a probability given by @xmath78 where @xmath79 is a constant chosen such that it normalizes the probability . for @xmath65
the model reduces to the viana - bray spin - glass model @xcite on erds - rnyi random graph .
we will refer to the two models defined above as model i ( [ i ] ) and model ii ( [ ii ] ) , respectively . both models should belong to the same universality class for the same value of @xmath63 , with respect to the critical and the low temperature properties . for
@xmath80 $ ] the models have a finite temperature spin - glass transition .
the critical point has a classical ( i.e. , gaussian ) character for @xmath81 $ ] , while it has a non - classical character for @xmath82 $ ] . as observed in @xcite , diluted models as the one in ( [ ii ] ) are convenient in numerical simulations since the number of interactions for each spin does not grow with the system size .
spending an equal amount of computational effort we can therefore study much bigger sizes than for model i , hoping that the finite - size corrections are comparable in both models .
let us discuss some scenarios for the behavior of the pdf @xmath83 corresponding to different possible physical situations . the simplest physical situation is the paramagnetic state where long - range order is absent . in this case
we have a finite correlation length @xmath84 and for large @xmath85 one can expect @xmath5 and @xmath6 to be sums of @xmath86 independent terms . the resulting probability distribution of @xmath5 and @xmath6 is a product of two independent gaussians with variance proportional to @xmath87 . we can then consider a condensed phase with only two pure states . in this case , the space average value of the overlap can take two values @xmath88 .
small fluctuations will still be gaussian as in the paramagnetic case . due to the symmetry of the model , where groups of spins at a given level of the hierarchy are on the same foot
, one can expect that large fluctuations where @xmath89 and @xmath90 ( or vice versa ) will imply a free - energy cost @xmath91 and have a probability @xmath92 . the most interesting possibility is a spin - glass phase with rsb . in this case one
has a zero mode in the free - energy associated to the existence of a couple of states with overlaps taking values in a finite interval @xcite .
if one chooses @xmath93 one finds a broad distribution @xmath94 , which for large @xmath85 is close to the limiting distribution @xmath4 .
consider first a system in absence of interactions at the @xmath95-th level .
the two subsystems are independent and @xmath96 .
if the interaction is switched on , we can expect a free - energy cost @xmath97 .
the overlap interface exponent is just given by @xmath98 due to the fact that the total interaction strength _ squared _ between the two parts scales as @xmath99 .
the value of the exponent @xmath13 as well as the function @xmath48 can be computed supposing rsb at the level @xmath85 .
the detailed calculation is quite involved and is presented in the next section ; here we just give the net result , valid in the regime @xmath100 $ ] .
neglecting prefactors we could not compute , it reads : @xmath101 notice the appearance of the @xmath102 function @xmath4 in the exponent of ( [ main ] ) . from this formula
one can extract the conditional probability of the difference @xmath43 for a fixed value of the semi - sum @xmath47 : @xmath103 equation ( [ diff ] ) summarizes our prediction for the hierarchical model . for a fixed value of the sum
, the difference of the overlap is distributed according to the exponential of the cube , which is different from a naive gaussian guess .
the coefficient of the exponential is equal to the function @xmath4 which can be evaluated in independent measurements . in numerical simulations it might be complicated to collect sufficient statistics to condition @xmath50 to the value of @xmath104 .
one can then turn to the unconditional distribution . in the large deviation regime , where @xmath105 , this should be dominated for large @xmath85 and finite @xmath50 by the value of @xmath104 which maximizes ( [ diff ] ) . for functions @xmath4 as the one commonly met in spin - glass systems this is the value @xmath106 which is the largest possible value of the overlap in the thermodynamic limit .
the form ( [ diff ] ) suggests that the order of magnitude of the typical fluctuations is @xmath107 .
in this regime , the unconditional distribution of @xmath50 involves the convolution of ( [ diff ] ) with a presently unknown prefactor and can not be computed .
all the form ( [ diff ] ) tells us for the unconditioned distribution in this regime is that the finite volume distribution of @xmath50 admits the scaling form @xmath108 in order to compute the scaling function @xmath109 the knowledge of the prefactor in ( [ diff ] ) would be necessary .
we remark that the exponent in this scaling is a strong consequence of rsb theory , a naive guess would have suggested a gaussian distribution with scaling variable @xmath110 .
we analyze model i through the replica method . in order to compute the free - energy ,
it is natural to consider a recursion that relates the average partition function of @xmath111 replicas @xmath112 , @xmath113 , with fixed mutual overlaps @xmath114 . defining @xmath115 = e_j\left[\sum_{{\bf s}}\exp\left(-\sum_{a=1}^n h_k^{j}(s_1^a, ... ,s_{2^k}^a)\right)\prod_{a < b}^{1,n}\delta \left(q_{ab}-\frac{1}{2^k}\sum_{i=1}^{2^k } s_i^a s_i^b \right)\right]\!,\ ] ] where @xmath116 denotes the average over the disorder , we can write : @xmath115 = \exp\left ( \frac{\beta^2}{4}2^{2(1-\sigma)k } \tr { \bf q}^2 \right ) \int { \cal d } { \bf q}_1{\cal d } { \bf q}_2 z_{k-1 } [ { \bf q}_1]z_{k-1}[{\bf q}_2]\delta \left ( { \bf q}-\frac{{\bf q}_1+{\bf q}_2}{2 } \right)\!. \label{zeta}\ ] ] for integer @xmath111 this is an exact relation based on the independence of the hamiltonians of the sub - systems at level @xmath85 . in principle , the thermodynamical properties of the system are encoded in this recursion and in its analytic continuations for @xmath117 .
for example it can be used to set up an epsilon expansion for the calculation of the critical indexes for @xmath118 $ ] @xcite . in this paper
we just use ( [ zeta ] ) to study the probability distribution ( [ diff ] ) .
this can be done with the technique of constrained replicas , introduced and discussed at length in @xcite . in order to consider constrained free - energies for two replicas with fixed overlaps
one should fix some of the elements of the matrix @xmath119 @xmath31 to the values of the constraints .
writing @xmath120 the constraint reads @xmath121 for @xmath31 and @xmath122 . introducing the replica free - energy at level @xmath85 for fixed @xmath5 and @xmath6 , @xmath123=e^{-\beta 2^k f_k[{\bf q}|q_1,q_2]}$ ]
we see that one needs in principle to compute : @xmath124 } = & \int { \cal d } { \bf q}_1{\cal d } { \bf q}_2 e^{\left ( \frac{\beta^2}{4}2^{2(1-\sigma)k } \tr ( { \bf q}_1+{\bf q}_2)^2 \right ) } e^{-\beta 2^k ( f_{k}[{\bf q}_1]+f_k[{\bf q}_2 ] ) } \times \nonumber \\ & \times \prod_{a=1}^{n'}\delta(q^1_{a , a+n'}-q_1)\delta(q^2_{a , a+n'}-q_2 ) .
\label{schif}\end{aligned}\ ] ] this form suggests that for large @xmath85 , supposing the knowledge about @xmath125 $ ] , the integral over @xmath126 and @xmath127 can be performed by the saddle - point approximation .
notice that the interaction term is sub - extensive and scales as @xmath128 , while the partial free - energies scale as the volume @xmath57 .
the interaction term does therefore not contribute to the saddle point , and the maximization with respect to the matrices @xmath129 can be performed separately in each of the two sub - systems . the value of the interaction term at the saddle point determines the free - energy difference @xmath19 .
we now study the consequences of the hypothesis that there is rsb in the system .
specifically , we suppose that in absence of any constraints rsb is described by a continuous parisi function @xmath130 taking values between the two extremes @xmath131 and @xmath132 . in the constrained problem ,
correspondingly , each of the matrices @xmath129 is parametrized by two functions @xmath133 and @xmath134 with @xmath135 $ ] . as implied by the analysis in @xcite
the free - energy in each of the two sub - systems is then independent of @xmath136 , and the function @xmath133 and @xmath134 can be directly related to the function @xmath130 of the unconstrained system by the relations @xmath137 where @xmath138 is the value of @xmath139 such that @xmath140 .
we then see that the free - energy difference from the unconstrained case is entirely due to the interaction term , which can be evaluated using the saddle - point value of the matrices @xmath141 and @xmath142 : @xmath143\!\!.\ ] ] substitution of ( [ p2 ] ) leads to @xmath144 where without loss of generality we have supposed @xmath145 .
the first term is just the contribution that can be expected if @xmath40 .
together with the two subsystems free - energy it just gives the free - energy of the system at the level @xmath95 .
the second term is associated to the free - energy excess needed to impose @xmath146 . for small @xmath147
we can expand this last term and find @xmath148 where @xmath149 is the inverse function of @xmath130
. inserted in ( [ schif ] ) and identifying @xmath150 with @xmath4 as discussed in length in @xcite leads to ( [ diff ] ) @xmath151
in this section we discuss the results of numerical simulations used to test the behavior of the probability distribution of the overlap difference in the two subsystems . in order to deal with pdf s of a single variable we concentrated to the unconditional probability for which we have the theoretical prediction ( [ uncond ] ) .
we did not try to test the more detailed prediction ( [ diff ] ) which would need the numerical determination joint pdf s of two variables .
we have simulated model ii for @xmath152 , using parallel tempering to thermalize the system at low temperatures .
we have tested our theoretical predictions concentrating on two values of @xmath63 : @xmath153 which lies in the classical region @xmath154 , where the spin - glass transition is well described by mean - field theory , and @xmath155 which lies in the non - classical region @xmath156 , where the exponents are nontrivial .
we have characterized the critical point of the model following the procedure that we describe in the appendix .
we estimate the critical temperatures to be equal to @xmath157 for @xmath153 and @xmath158 for @xmath155 . to test the predictions described in the previous sections we measured the distribution of the variables @xmath104 and @xmath50 as a function of temperature and system size .
we considered systems sizes of @xmath159 and @xmath160 spins .
averages were performed over 4000 samples for the smaller systems and 700 samples for the largest system .
the configurations are thermalized during the first @xmath161 monte carlo ( mc ) sweeps of the runs and then data is collected for the subsequent @xmath161 mc steps .
the first prediction we test is the validity of the cube - exponential form in the tails for large @xmath50 .
this is well observed in all our simulations .
a typical example of our findings is depicted in figure ( [ fig1 ] ) where we plot the function @xmath162 for @xmath163 for the two different values of @xmath63 in the classical and non - classical regime , respectively .
the data are plotted together with functions of form @xmath164 which should be considered as a guide to the eye rather than the best fit .
the parameters @xmath165 and @xmath166 were fixed by eye to be equal to @xmath167 , @xmath168 for @xmath155 and @xmath169 , @xmath170 for @xmath153 .
a best fit procedure results to be sensitive to the chosen fitting interval and to the tails of the distribution that represent probabilities too small to be correctly estimated with our statistics . despite these caveats we believe that our data give an indication in favor of the cubic behavior for both values of @xmath63 .
we then investigated the scaling of the pdf of @xmath50 with the system size , contrasting it with the behavior of the pdf of @xmath104 . in figures
[ fig2 ] and [ fig4 ] we display the function @xmath171 for the two values of @xmath63 at low temperatures .
the function @xmath171 has the characteristic appearance of the one of systems developing rsb for large volumes , with two symmetric peaks and a non - zero part for @xmath172 .
figures [ fig3 ] and [ fig5 ] in contrast show that the distribution of @xmath50 , @xmath162 , is unimodal around zero .
its width is as expected a decreasing function of @xmath85 . in the insets
the unscaled data are presented , while the in the main panel the result of scaling the data using the variable @xmath173 is shown .
we judge the scaling ( [ uncond ] ) to be very satisfactory , though not perfect for the values of @xmath85 we explored .
indeed , while for @xmath153 we could not find a the value of the ratio @xmath174 producing a better data collapse than the theoretical value @xmath175 , for @xmath155 the value @xmath176 produces a better data collapse than the theoretical value @xmath177 .
we believe that this discrepancy is due to finite size effects , but we can not exclude at present that the theory should be amended in the non - classical region .
our data show that the best fitting exponent is largely temperature independent in the low temperature region .
a direct scaling of the data in the high temperature region for the values of @xmath178 that we dispose , produces an effective exponent that crosses over slowly from the low temperature value towards the paramagnetic value @xmath179 .
we believe that we are seeing a preasymptotic behavior due to the influence of the critical fixed point . this influence could be particularly marked due to the power - law interactions where the critical fixed point continues to attract the system on relatively large scales .
larger and larger values of @xmath85 are necessary to observe the paramagnetic behavior closer and closer to the critical temperature .
in this paper we have proposed a new method to determine numerically the overlap interface exponents first defined in @xcite .
these can be obtained by looking at the finite - size scaling of the pdf of the differences of the overlaps between two replicas in two subsystems .
we have applied the definition to hierarchical models , where we could give a theoretical prediction for the scaling of the overlap differences with size .
though the interface exponent is naturally dictated by the model , the dependence in @xmath50 is found to be nontrivial in presence of rsb .
we tested this dependence in numerical simulations finding very satisfactory agreement both in the classical and in the non - classical region .
these results confirm the interest of hierarchical spin - glass models , that combine analytical tractability , nontrivial critical points and rsb low temperature phases .
* acknowledgments * it is a pleasure to thank o. c. martin , m.mzard , f. ricci - tersenghi , p. contucci , c. giardin , c. giberti and c. vernia for interesting discussions .
in this appendix we discuss the characterization of the critical point of the model for the values of the interaction parameter @xmath63 that we have considered in the text .
the model has been simulated with the parallel tempering algorithm , with 10 values of the temperature , using @xmath161 thermalization steps before collecting data in the @xmath161 steps .
we considered systems sizes of @xmath180 and @xmath160 spins .
averages were performed over 4000 samples for the smaller systems and 700 samples for the largest system . simulating two replicas in parallel ,
we have studied the second and the fourth moment of the distribution of the mutual overlap @xmath181 and @xmath182 .
we have identified the critical temperature and the exponents @xmath183 and @xmath184 using finite size scaling through the behavior of @xmath6 , @xmath185 and the corresponding binder parameter @xcite @xmath186 . in the non - classical region , @xmath187 where finite - size scaling should hold the various parameters exhibit the following dependence on the temperature and system size @xmath188 : @xmath189 the exponent @xmath183 should not renormalize in long range models , and analogously to the euclidian 1d model take the value @xmath190 both in the non - classical and in the classical regions@xcite . in the classical region , @xmath154 , the scaling implied by ( [ noncl ] )
does not hold @xcite .
it is possible to show , using the fact that the critical theory is described by a cubic action analogous to the one for short - range spin glasses @xcite , that the various quantities scale according to the following : @xmath191 the exponent @xmath192 in the scaling functions can be derived from dimensional analysis from the cubic action .
the exponent @xmath193 takes the value @xmath194 independently of @xmath63 as can be checked from the scaling relation @xmath195 with @xmath196 .
let us now turn to the data considering the non - classical region first .
to analyze the data we observe the following procedure : we first estimate the critical temperature from the crossing point of the binder parameter and the rescaled values , especially , @xmath197 and @xmath198 since these provide for a cleaner crossing than the binder parameter .
we then fix the value of @xmath184 in order to collapse the curves .
the result for @xmath155 is shown in figure [ fig6 ] , we present the data for @xmath6 , @xmath185 and @xmath199 scaled as in ( [ sc ] ) using a value of @xmath200 . if we try to use the scaling ( [ noncl ] ) in the classical regime we get inconsistent results : although we obtain an approximate crossing of the curves for the three quantities , the temperatures at which the curves cross do clearly not coincide .
the crossing of the binder parameter indicates @xmath201 and as shown in figure [ fig6 ] for @xmath153 , we obtain consistent scaling assuming the form ( [ sc ] ) and not ( [ noncl ] ) . | we discuss interfaces in spin glasses .
we present new theoretical results and a numerical method to characterize overlap interfaces and the stability of the spin - glass phase in extended disordered systems .
we use this definition to characterize the low temperature phase of hierarchical spin - glass models .
we use the replica symmetry breaking theory to evaluate the cost for an overlap interface , which in these models is particularly simple .
a comparison of our results from numerical simulations with the theoretical predictions shows good agreement . |
in this paper we report _ ab initio _ calculations of equilibrium geometries , formation energies , and extrinsic gap levels of vacancies in @xmath0-quartz .
we concentrate on oxygen vacancies , which are common defects in both silica ( amorphous sio@xmath1 ) and quartz , and alleged to be precursors of other more complex defects @xcite . in view of the short - range order conservation upon thermal amorphization of quartz , which basically rules out complex multicenter native defects ,
the state - of - the - art results for oxygen vacancies in crystalline quartz to be discussed below are expected to shed some light on defect phenomenology in amorphous silica .
recent ab initio molecular dynamics simulations of the amorphization of a small ( 128-atom ) sample of @xmath0-quartz via melt quenching @xcite have given evidence for an almost perfect conservation of short range order in the amorphous phase ; the structure obtained in the simulation @xcite maintains chemical order , with si - centered tetrahedra linked by corner - sharing oxygens . in these simulations , disorder manifests itself merely as a broadening of the distribution of si - o - si angle between linked tetrahedra ; the o - si - o angle distribution has instead a sharp maximum close to the tetrahedral angle of 109@xmath6 .
edge - sharing tetrahedra are absent in the amorphous phase , but they occasionally appear in the liquid @xcite . on the other hand , even in the liquid
the dominant defects are simple miscoordinated atoms ( 3- or 5-fold coordinated si , and 1- or 2-fold coordinated o ) , occurring with small ( @xmath7 10 % ) probability . although the predictions of ab initio molecular dynamics may be considered too optimistic ( mostly , because of the smallness of the simulated amorphous sample ) , it is clear that even the simplest defects are not easily formed thermally in sio@xmath1 .
more complex defects are likely to be even more rare .
a study of lattice vacancies seems therefore timely , also because , although many semiempirical studies @xcite have appeared in the past , even these simple defects have not been widely studied at the ab initio level ( recent density - functional - theory studies are reported in @xcite and @xcite ) .
in particular , no systematic study exists , to our knowledge , dealing with the charge states of the o vacancy beyond the neutral and + 1 states @xcite .
we perform ab initio density - functional theory ( dft)@xcite plane - waves pseudopotential calculations , using ultrasoft - pseudopotentials @xcite , which guarantee a high degree of transferability and a fast convergence in fourier space of all the relevant quantities @xcite .
a plane - waves basis is used with a cut - off at 20 ryd .
a conjugate - gradient total - energy minimization @xcite is used to obtain the self - consistent electronic ground - state at fixed ions , and an efficient relaxation method @xcite is used to reach the minimum energy structure following hellmann - feynman forces .
we employ 36-atom orthorombic supercells with no imposed symmetry , and one special * k * point as brillouin zone sampling .
this setting gives structural properties in agreement with those obtained by liu _
@xcite , matching experiment to better than 1 % , apart from two o internal parameters having a deviation of @xmath7 2% .
the formation energy for an o vacancy in sio@xmath1 in charge state _ q _ ( where , conventionally , @xmath2 is the number of electrons transferred from the defect to a reservoir of chemical potential @xmath8 ) is given by @xmath9 where @xmath10 is the total energy difference between the defected supercell in charge state @xmath2 and the perfect bulk supercell , and @xmath8 is the electron chemical potential ( i.e the fermi level in our t=0 calculation ) .
the zero of @xmath8 may be assumed to be the top valence band energy @xmath11 of the defected crystal in the charge state @xmath2 .
the chemical potential @xmath12 is one half of the total energy of the spin - unpolarized o@xmath1 molecule ( calculated in a large supercell ) .
as expressed above , the formation energy depends on the charge state and on the fermi level .
the latter embodies the effects of background doping , competing defect species , or externally injected carriers .
the eigenvalues of the hamiltonian of a crystal with a fixed number of electrons have no rigorous relationship to observable optical transition , unless they are many - body quasiparticle energies .
this is also the case for defected crystals , and for transitions involving extrinsic levels therein . in particular , transitions to and from extrinsic levels
can not simply be extracted as differences of dft eigenvalues of e.g. the neutral state of the defect .
this is true for any one - body theory , including hartree - fock calculations @xcite , since koopman s theorem gives ( e.g. ) an ionization potential which is not exact even in principle . within dft
one can resort to total energy differences , since the dft total energy is exact in principle , apart from the local density approximation ( lda ) and other technical issues of practical relevance @xcite .
in particular , in a dft calculation one calculates formation ( i.e. total ) energy differences between the relevant systems . for e.g.
the case of @xmath2=0 and @xmath2=1 , this difference is , by definition , the energy needed to extract an electron from the @xmath2=1 state ( ionization of the charged center ) leaving behind a @xmath2=0 center , or to add one to the @xmath2=0 state ( binding to the neutral center ) transforming it into a @xmath2=1 state .
when account for geometrical changes upon occupation is taken , this difference is the thermal excitation level ; `` thermal '' indicates that the structure is allowed sufficient time to rearrange to its new equilibrium position .
we concentrate on the calculation of this quantity , and discuss later optical transitions .
apart from basic choices such as the use of the lda as opposed to gradient - corrected functionals , the uncertainties in the predicted defect energy levels arise mainly from the alignment of the energy zero of differently - charged supercells .
since we choose conventionally the fermi level zero at the valence band top , @xmath13 0 charge states have a constant energy offset @xmath14 away from the neutral state energy ( eq . [ eform ] ) .
we proceed by calculating the _ bulk _ valence band top @xmath15 in a bulk supercell , and express the valence top in the charged supercell as @xmath16 where @xmath17 is the difference of the macroscopic averages@xcite of the local potentials of the charged supercell and of the bulk supercell , @xmath18 this quantity is evaluated in an appropriate bulk - like region , i.e. regions of the simulation cell where the local potential is unaffected by the defect .
the resulting uncertainties are at most of order @xmath19 0.1 ev in the thermal ionization levels .
for the neutral state of the vacancy , the formation energy is 6.97 ev .
such large energy is in accord with the need for irradiation to produce vacancies in the first place . as seen from fig .
[ f1eform ] , the vacancy is stable in the @xmath2=+3 state for fermi level values of up to 2 ev above the valence band top .
thereafter , the neutral state becomes stable , and remains such for fermi levels up to about 5.5 ev .
the neutral state is therefore the ground state of the oxygen vacancy in nominally undoped quartz , i.e. for the fermi level at mid - gap ( @xmath7 4.5 ev ) . increasing further the fermi level , the vacancy becomes negatively charged .
the @xmath20 charge state is the most stable state for fermi levels ranging from about 5.5 ev to 8 ev ( neglecting the small fermi level range from 5.5 ev to 5.7 ev , where the 2 state prevails ) . while the thermal level sequence which is usually expected is @xmath2
= + 3 , + 2 , + 1 , 0 , 1 , 2 , 3 ... , in the present case we find a negative - u behavior @xcite between the triply positive state and the neutral state , and between the neutral and the @xmath20 state : the @xmath3 + 2 and @xmath3 + 1 states are unstable against the release of a further electron , and the @xmath2=1 and @xmath2=2 states are unstable towards the capture of a further electron .
this is due to a strong electron - lattice coupling , as is discussed below .
the realization of large-@xmath2 negative and positive states is not unusual for wide gap materials .
the @xmath21 vacancy is a symmetric combination of si dangling bonds .
the si neighboring to the vacancy relax almost symmetrically , by 9.5 % of the unrelaxed si - o bond length , towards the vacant site . in fig .
[ f2rho0 ] , we display density isosurfaces for the defect state of the neutral vacancy .
the only stable positive charge state is the @xmath2=+3 state .
the two si atoms neighboring to the vacancy relax strongly away from their original positions in an asymmetric fashion : their distance to the vacancy site increases from 1.62 to 2.16 , and from 1.61 to 2.81 , respectively .
the two si atom are three - fold coordinated and close to co - planar with the neighboring oxygens .
the charge density of the last occupied state is delocalized , suggesting that the defect level is resonant with the valence band .
the thermodynamical instability of the singly positive vacancy signals potential difficulties in previously proposed models @xcite of the so called e@xmath22 center .
the @xmath23 charge state is the antibonding combination of the same dangling bonds as above , as can be seen in fig .
[ f3rho-1 ] ( compare with the previous figure ) .
additional relaxations with respect to the neutral state are below 1% .
thus , electronic promotion into this state will require a ( rather large ) energy of almost purely electronic origin .
in the @xmath24 charge state , the first antibonding state gets completely filled .
in contrast to the @xmath23 , a large , markedly asymmetric relaxation occurs .
this can be seen in the structure model of fig .
[ fxmol ] . in this geometry atom si@xmath25 has moved towards the vacancy by about 20 % of the pristine unrelaxed bond length , whereas si@xmath1 has moved away from the vacancy by about 3 % .
the tetrahedron centered on si@xmath25 remains essentially undistorted , while that centered on si@xmath1 is distorded concurrently with the neighboring tetrahedron centered on a third si atom , si@xmath26 .
this effect is due to a large displacement of the bridging o atom . due to the large relaxation energy gain and to the associated electronic rehybridization ( visible in fig .
[ f4rho-2 ] , where the relevant charge density isosurfaces are displayed ) , the @xmath24 state is lower in energy than the @xmath23 ( fig.[f1eform ] ) over most of the e@xmath27 range , despite the electron - electron repulsion .
thus a negative - u effect takes place .
the 2 state becomes favored over the neutral at a fermi level of 5.5 ev . at a fermi level about 0.2 ev
higher the @xmath20 state charge becomes favored .
si@xmath25 moves towards the vacancy site by an additional 7 % ( its distance from the vacancy is about 30 % shorther than the unrelaxed bond length ) , while si@xmath1 moves closer by 2 % ( its distance from the vacancy is about 1 % longer than the unrelaxed bond length ) .
the stability of the @xmath20 state over the @xmath24 state for a wide range of e@xmath27 is somewhat puzzling in view of the large repulsion to be expected among electrons in the center .
the prominent feature of this state is indeed that the additional electron minimizes the coulomb repulsion with the preextant ones by occupying a state in the interstice between the vacancy - adjacent si@xmath25 and si@xmath1 atoms , and the si@xmath26 atom belonging to the tetrahedron neighboring to si@xmath1 and closest to the vacancy .
the distance between si@xmath1 and si@xmath26 is reduced to only 4.3 a.u .
, to be compared with a distance of 5.8 a.u . in the unrelaxed structure .
this is due to the aforementioned distortion of the si@xmath1-o - si@xmath26 bridge linking the two tetrahedral units , whereby the corresponding angle decreases from the usual 144@xmath6 to only about 80@xmath6 .
while this distortion is in fact already qualitatively present in the @xmath24 charge state , the additional electron in the @xmath20 state is effectively bound by si@xmath26 , as can be seen from fig .
[ f5rho-3 ] . in view of its electronic structure
, the 3 charge state may be a candidate e@xmath22-like center ( see below ) .
the @xmath20 state is stable up to extreme @xmath5-type conditions ( fermi level @xmath7 8.1 ev ) , whereby the @xmath28 state sets in .
the fourth electron fills in the state whose occupation was initiated with q=3 , with marginal additional relaxation .
the charge density and geometry are very close to the q=3 . by the time one gets to extreme @xmath4 or @xmath5-type conditions ,
the formation energies of the vacancy become very small .
vacancy - driven compensation of acceptor and donor impurities will be very efficient , and we expect that @xmath5-type doping of quartz will only be achieved with extreme difficulty , if at all .
the franck - condon ( fc ) shift @xmath29@xmath30 is defined as the total energy difference between the system in charge state @xmath2 in its own equilibrium configuration and the same system in the equilibrium configuration of charge state @xmath31 : @xmath32 - e_{\rm tot}^q [ { \bf r}_{q}]\ , .\ ] ] the fc shift is symmetric upon exchange of @xmath2 and @xmath31 if the dependence of the total energy on configurational coordinates is quadratic and with identical force constants in the two charge states .
however , given the large distortions involved in the present case , the fc shift might be asymmetric ( expecially for large shifts ) both because of anharmonicity and of force - constant differences .
these expectations are indeed verified in our calculations .
we find large asymmetries for large shifts , which are consistent with force constants differing by factors of up to about @xmath33 assuming a single quadratic configurational cooordinate .
we express for convenience the fc shifts between the non - positive ( non - negative ) charge states as a matrix @xmath34 ( @xmath35 ) whose elements @xmath29@xmath36 ( @xmath29@xmath37 ) are the fc shifts for the @xmath38 ( @xmath39 ) state in the geometry of the @xmath40 ( @xmath41 ) state ; this matrix would be symmetric in the symmetric - fc hypotesis , and its diagonal elements are null . for negative states , with @xmath2 between 0 and 3 , we have @xmath42 while for positively charged states ( @xmath2 between 0 and + 3 ) we have @xmath43 all energies being expressed in ev .
note that only the @xmath29@xmath44 elements are needed in the discussion to follow .
the optical transition energy is the total energy difference of the two ( differently charged ) initial and final systems in a fixed geometry . alternatively
( but equivalently ) , optical levels can be expressed as combinations of thermal levels and fc shifts .
we only consider one - electron ( i.e. first - order ) processes .
this kind of process amounts to transferring only one electron at a time into , or out of , a defect level .
the transitions we consider are absorption ( an electron is transferred from the valence states reservoir into the defect , or transferred from the defect to the conduction band via an internal excitation of e.g. the 3 state into the 2 state ) , recombination ( the electron drops into the defect level from a reservoir of e.g. optically - excited conduction electrons ) , and single - particle deexcitation of a metastable charged center by recombination of an electron into the valence band .
we assume that all processes involved , e.g. successive absorption transitions ( @xmath45 , @xmath46 , @xmath47 ... ) , take place on time scales shorter than those of lattice vibrations , so that ( e.g. ) absorption occurs in a frozen structure .
this assumption seems plausible in view of typical phonon frequencies and optical transition times , especially in large excitation densities .
we therefore assume that the structure remain frozen in the equilibrium configuration of a given state @xmath48 ( typically the initial state ) throughout the transition cascade .
if the ground state of the defect is charge state @xmath2 , in terms of thermal levels @xmath49 $ ] and fc shifts @xmath29@xmath30 the optical absorption energy for electronic transitions from the valence band into charge state @xmath50 is @xmath51 + { \cal{f } } _ { q-1,q_f } - { \cal{f } } _ { q , q_f}.\ ] ] for internal absorption excitations ( transforming e.g. @xmath24 into @xmath23 by electron excitation in the conduction band ) we have instead @xmath52 + { \cal{f } } _ { q , q_f } - { \cal{f } } _ { q-1,q_f}.\ ] ] within the same assumptions , for the conduction band - to - acceptor recombination we have _ emissions _ at @xmath53 with @xmath54 ev , the experimental gap @xcite of @xmath55sio@xmath1 .
once again we implicitly assumed that subsequent recombinations are instantaneous on the scale of ionic relaxation .
deexcitation is slightly more complicated .
suppose that the vacancies have been occupied with a number of electrons e.g. between 1 and 4 .
assume further that the lifetimes of each charge states are larger than the inverse of typical phonon frequencies ( @xmath7 10 thz ) .
this assumption seems again admissible in view of typical decay constants of observed luminescence bands in silica @xcite being of order 10@xmath56 to 10@xmath57 sec .
then , each charge state can relax to its own ground state before deexcitation . for each vacancy in a given charge state ,
a cascade of deexcitation processes occurs , whereby each charge state ( between the initial one , say @xmath58 , to @xmath21 ) is occupied in sequence , and it loses one electron to the valence band while in _ its own _ equilibrium geometry . expressing again total - energy differences in terms of thermal levels and fc shifts , we obtain the dexcitation emission energies from charge state @xmath2 as @xmath59 - { \cal{f } } _ { q , q-1}.\ ] ] for internal deexcitation the formula reads @xmath60 + { \cal{f } } _ { q-1,q}.\ ] ] here
we did not consider selection rules possibly preventing some of the transitions involved , and remain with an open question on this issue .
as mentioned above , quantitative predictions depend on which ground state is actually realized .
this makes the optical properties of this system quite rich and complex . in table
i we summarize the main groups of absorption and emission transitions involving different ground states .
if the fermi level is at mid - gap ( i.e. the solid is undoped ) , the ground state is the neutral state . in moderate @xmath5-type conditions
, the ground state will typically be the 3 state .
the relevant thermal levels are @xmath61 $ ] = 6.24 ev , @xmath62 $ ] = 5.47 ev , and @xmath63 $ ] = 5.82 ev .
the fc shifts have been given previously .
in @xmath4-type conditions , the ground state will be the + 3 state .
the relevant thermal levels are @xmath64 $ ] = 2.48 ev , @xmath65 $ ] = 2.69 ev , and @xmath66 $ ] = 0.72 ev ( the fc shifts have been given previously ) . for the neutral state , the absorption band
is centered around 6.6 ev and the emission band at about 2.3 ev .
for the 2 charge state there is an absorption band at about 6.3 ev , while the emission one is centered around 3.0 ev . for the @xmath20 state
, the absorption bands are located around 3.2 ev , 6.0 ev , and 7.1 ev .
these correspond to internal excitations @xmath67 , @xmath68 , and @xmath69 respectively . while calculated line intensities are not available , we speculate that the latter components may give rise to a band centered around 6.3 ev , close to an experimentally observed absorption band ( 6.2 ev ) usually assumed to be related to the e@xmath22 center . on the other hand ,
the emission bands are around 2.7 ev , close to the 2.7/2.8 ev emission band generally attributed to the oxygen vacancy , and at 3.2 ev , where a band is also observed experimentally in silica . for the @xmath70 state , absorptions
( either normal or internal ) are predicted at 2.2 ev , 3.1 ev and 5.0 ev , while emissions are at 1.9 ev and 0.9 ev .
we note at this point that the triply - negative charge state of the oxygen vacancy exhibits several of the characteristic features of the e@xmath22 center @xcite in @xmath0-quartz .
both the transition generating the center ( if the starting ground state is the neutral vacancy ) and that characteristic of the center itself are correctly predicted around 6 ev .
while such transition energies appear also in the predicted spectra of other possible ground states , only the electronic structure of the @xmath2=3 state seems compatible with hyperfine - field and epr measurements @xcite .
these experiments indicate the presence of two weak hyperfine interactions and a strong one associated with the oxygen vacancy . in our 3 state the weak signals
can be attributed to si@xmath25 and si@xmath1 , while the strong signal comes from the unpaired electron prevailingly bound to si@xmath26 .
such attributions could not be made for the @xmath2=+3 , 0 , and 2 ground states .
the unpaired electron lobe is approximately directed along the direction of the `` short bond '' @xcite in the ideal structure , and as can be seen from fig .
[ fxmol ] , si@xmath26 is close to three fold coordinated , as it is generally accepted that e@xmath22 should be @xcite .
the main objection to identifying the @xmath2=3 state of the vacancy with e@xmath22 comes from experiments showing that e@xmath22 is positively charged in mos si - sio@xmath1 junctions @xcite .
however , this result is not always reproduced in other instances @xcite .
it would be interesting to analyze further these experiments in the light of our results , taking into account for instance band offset effects at the si / sio@xmath1 interface .
as to magnetic data , recent calculations @xcite of the hyperfine tensor of a distorted variant of the singly positive vacancy have resulted in good agreement with the data pertaining to the e@xmath22 center @xcite , giving support to the attribution of e@xmath22 to a positive charge state of the vacancy . on the other hand
our results indicate that the @xmath2=+1 vacancy is unstable towards becoming triply charged , and that the @xmath2=+3 state appears to be a valence resonance , lacking features which could explain the peculiar magnetic signature of e@xmath22 .
it is clear that additional work is needed to resolve this puzzle .
nevertheless , the 3 oxygen vacancy seems to carry several e@xmath22-like features , and it may be considered as a possible new variety of e@xmath22 defect .
we have presented ab initio results on oxygen vacancies in @xmath55sio@xmath1 .
the ground states for the vacancy as function of background doping are the triply - charged positive , neutral , and triply - charged negative states .
the high - charge states are stabilized by large , asymmetric distortions involving at least three tetrahedral units neighboring to the vacancy . in connection with the possible different ground states we predict optical absorption energies in the 3.2 ev and 6 - 7 ev ranges , and emissions in the 2.3 - 3.2 ev range .
based on predicted optical transitions , geometry , and electronic structure , we suggested a possible new e@xmath22-like center , namely the triply - charged negative oxygen vacancy .
we thank david vanderbilt for his codes ( coauthored in part by r. d. king - smith ) , used in this work in a modified version .
we thank fabio bernardini for useful discussions .
this work was supported in part by cineca bologna through supercomputing grant 95 - 1506 . c. m. c. was supported by the istituto nazionale di fisica della materia within progetto sud , funded by the european union .
exchange of correspondence with a. pasquarello provided usefuls insights .
d. l. griscom , j. non - cryst .
solids * 73 * , 5 ( 1985 ) .
j. sarnthein , a. pasquarello , and r. car , phys .
lett . * 74 * , 4682 ( 1995 ) .
see e.g. j. k. rudra and w. beall fowler , phys .
rev b * 35 * , 8223 ( 1987 ) for an entry to the literature .
d. c. allan and m. p. teter , j. am .
. soc . * 73 * , 3247 ( 1990 ) .
m. boero , a. pasquarello , j. sarnthein , and r. car , phys .
* 78 * , 887 ( 1997 ) .
r. dreizler and e. k. u. gross , _ density functional theory _ , ( springer , berlin , 1990 ) .
exchange - correlation energy by d. m. ceperley and b. j. alder , phys .
* 45 * , 566 ( 1980 ) as parametrized by j. p. perdew and a. zunger , phys .
b * 23 * , 5048 ( 1981 ) .
d. vanderbilt , phys .
b * 41 * , 7892 ( 1990 ) .
r. d. king - smith and d. vanderbilt , phys .
b * 49 * , 5828 ( 1994 ) .
f. bernardini , to be published .
f. liu , s. garofalini , r. d. king - smith , and d. vanderbilt , phys .
b * 49 * , 12528 ( 1994 ) . k. c. snyder and w. beall fowler , phys .
b * 48 * , 13238 ( 1993 ) .
see e.g. w. e. pickett , comp .
phys . rep .
* 9 * , 115 ( 1989 ) a. baldereschi .
s. baroni , and r. resta , phys .
* 61 * , 734 ( 1988 ) .
u. scherz and m. scheffler , in _ semiconductors and semimetals _ , vol .
38 , e. weber ed .
( academic press , new york 1992 ) , p.1 . t. h. distefano and d. e. eastman , solid state commun .
* 9 * , 2259 ( 1970 ) .
see e.g. h. nishikawa _ et al .
rev b * 45 * , 586 ( 1992 ) .
r. a. weeks , j. appl .
* 27 * , 1376 ( 1956 ) ; r. h. silsbee , _ ibid . *
32 * _ , 1459 ( 1961 ) ; l. e. halliburton , m. g. jani , and r. b. bossoli nucl .
instr . meth .
b1 , 192 ( 1984 ) .
g. griscom , nucl .
instr . meth .
b*1 * , 481 ( 1984 ) .
w. l. warren , p. m. lenahan , and c. j. brinker , j. non - cryst .
solids * 136 * , 151 ( 1991 ) and reference therein .
[ 52 ] to [ 56 ] of our ref . | extrinsic levels , formation energies , and relaxation geometries are calculated ab initio for oxygen vacancies in @xmath0-quartz sio@xmath1 .
the vacancy is found to be thermodynamically stable in the charge states @xmath2=+3 , @xmath2=0 , @xmath32 , and @xmath2=3 .
the charged states are stabilized by large and asymmetric distortions near the vacancy site .
concurrently , franck - condon shifts for absorption and recombination related to these states are found to be strongly asymmetric . in undoped quartz , the ground state of the vacancy is the neutral charge state , while for moderate @xmath4-type and @xmath5-type doping , the + 3 and 3 states are favored , respectively , over a wide fermi level window .
optical transitions related to the vacancy are predicted at around 3 ev and 6.5 ev ( absorption ) and 2.5 to 3.0 ev ( emission ) , depending on the charge state of the ground state . |
many unusual features of the high temperature superconductors ( htsc ) are understood in terms of a momentum dependent order parameter @xmath1 .
it follows from the success of these developments naturally , that investigating the properties of unconventional condensates in the density wave sector may as well turn out to be a fruitful enterprise . indeed , in the last few years researchers encountered mysterious low temperature phases in a number of materials , where a clear and robust thermodynamic phase transition is not accompanied by an order parameter detectable by conventional means
this situation is often referred to as hidden order .
prime example is the 17.5 k transition in uru@xmath0si@xmath0 with very small magnetic moment in the low temperature phase .
it has been suggested@xcite , that an unconventional spin density wave ( usdw ) may be responsible for this behavior , since in the clean system usdw does not exhibit periodic spin density modulation due to the vanishing momentum average of the order parameter .
the origin of the small but finite magnetic moment , termed micromagnetism , however remains elusive .
the magnetic phase diagram of some @xmath2-(et)@xmath0 salts also offers a low temperature phase ( ltp ) with resemblance to charge density waves ( cdw)@xcite , but without the characteristic x - ray satellites .
in this situation a quasi - one dimensional ucdw scenario@xcite is called for , and it s case is strengthened by the evaluation of the corresponding threshold electric field for sliding conductivity@xcite .
finally , the pseudogap phase of high temperature superconductors was proposed to be a kind of ucdw as well@xcite .
however , micromagnetism was also observed in ybco by neutron scattering@xcite . in this paper
we propose a mechanism by which the usdw exhibits weak priodic spin density modulation in the presence of random magnetic impurities . comparing the temperature dependence of neutron scattering intensity due to this modulation with experimental data on uru@xmath0si@xmath0@xcite , we argue that this mechanism is responsible for the micromagnetism in this material , and possibly also in the pseudogap phase of htsc .
the unconventional spin density wave ( usdw ) with spin polarization in the @xmath2 direction ( @xmath3 ) is described in the nambu formalism by the green s function @xmath4 , where the @xmath5 and @xmath6 pauli matrices operate on the electron - hole , and spin space respectively , @xmath7 is the electron spectrum measured from the fermi energy , and @xmath1 is the momentum dependent order parameter . in the quasi - one dimensional case@xcite it can be for example @xmath8 ,
while for uru@xmath0si@xmath0 @xmath9 $ ] has been proposed@xcite . clearly , due to the vanishing momentum average of @xmath1 , the expectation value of any component of the spin density is zero@xcite .
it turns out , that the quantity which does have a periodic modulation with the nesting vector @xmath10 , is the spin current density@xcite .
nevertheless , in the ideal system there is no spin density modulation , therefore the micromagnetism remains unexplained , although the green s function definitely prefers a unique direction ( @xmath2 ) in spin space .
there is no such spin direction for ucdw , therefore the mechanism for micromagnetism proposed below will not work in that case .
we assume that we have magnetic impurities in our sample at positions @xmath11 with spin @xmath12 interacting with the electron spin density through a heisenberg exchange coupling @xmath13 .
the first diagram on fig .
[ fig : graph ] shows the change in the thermodynamic potential due to impurities : @xmath14,\label{tdpot}\ ] ] where @xmath15 is the matrix element of the impurity potential between bloch states ( @xmath16 spanning the electron hole space ) .
@xmath17 vanishes for any kind of cdw , but for usdw each impurity feels a potential energy @xmath18 , trying to align the inpurity spin parallel to the spin direction favoured by the usdw , including its periodicity given by @xmath10 .
@xmath19 is the component of the interaction driving the usdw transition , while @xmath20 is the coefficient of the term originating from @xmath21 ( a matrix element of the exchange integral @xmath22 between bloch states ) , having the same momentum dependence as the order paprameter does ( for example @xmath23 ) .
uru once the impurity spins are coherently aligned much like in a conventional sdw , the electron spin polarization around them will also develop coherently , facilitating the observation of a magnetic order by neutron scattering .
the electron spin density induced by impurities can be calculated from the second diagram on fig . [
fig : graph ] as @xmath24 , \label{spin}\ ] ] where @xmath25_{s^\prime , s}= \varphi^*_{{\bf k}^\prime+s^\prime{\bf q}/2}({\bf r}){\mbox{\boldmath$\sigma$ } } \varphi_{{\bf k}+s{\bf q}/2}({\bf r})$ ] .
a similar expression has already been evaluated@xcite for the impurity pinning potential in udw , which allows us to express the electron spin density at the impurity site @xmath11 .
the excess spin polarization due to the usdw turns out to be parallel to the usdw s preferred spin direction ( @xmath2 ) as well , and the temperature dependence of its magnitude normalized to the @xmath26 value is given by @xmath27 the neutron scattering intensity is proportional to the square of this amplitude , and its normalized temperature dependence is plotted on fig . [
fig : urusi ] , along with the experimental data@xcite on uru@xmath0si@xmath0 .
[ t][b][1][0]@xmath28 [ b][t][1][0]@xmath29
we have proposed a mechanism for the development of micromagnetism in usdw , and shown that the temperature dependence of the magnetic moment is consistent with measurements on uru@xmath0si@xmath0 , lending further support for the suggestion that the low temperature phase of this material is a usdw .
we also believe that the same mechanism may apply for the pseudogap phase in htsc@xcite , suggesting that it is not ucdw , but usdw instead .
this is further corroborated by the optical dichroism observed in this phase@xcite .
we thank b. keimer and h. amitsuka for stimulating discussions .
this work was supported by the hungarian scientific research fund under grant numbers otka t032162 and t029877 , and by the ministry of education under grant number fkfp 0029/1999 . | it has been proposed , that unconventional density waves ( udw ) are possible candidates for systems with hidden order parameter . unlike in conventional density waves ,
no periodic modulation of either the charge- , or the spin - density is present in udw , in spite of a clear thermodynamic signal .
although the unconventional spin density wave ( usdw ) has been suggested for the `` antiferromagnetic '' phase of uru@xmath0si@xmath0 , the micromagnetism seen by neutron scattering has not been understood .
we present here the calculation of the local spin density due to impurities in usdw , which describes quantitatively the neutron scattering data by amitsuka _ et.al._. further , we propose that the pseudogap phase in high temperature superconductors ( htsc ) should also be usdw . strong evidence for this are the micromagnetism seen by sidis _
et.al._ , and the optical dichroism seen by campuzano _ |
in 1969 stuart kauffman started to study random boolean networks as simple models of genetic regulatory networks @xcite .
random boolean networks that consists of a set of boolean _ gates _ that are capable of storing a single boolean value . at discrete time
steps these gates store a new value according to an initially chosen random boolean function , which receives its inputs from random chosen gates .
we will give a more formal definition later .
kauffman made numerical studies of random networks , where the functions are chosen from the set of all boolean functions with @xmath0 arguments ( the so called _
@xmath2-networks _ ) .
he recognised that if @xmath3 , the random networks exhibit a remarkable form of ordered behaviour : the limit cycles are small , the number of _ ineffective gates _ , which are gates that can be perturbed without changing the asymptotic behaviour , and the number of _ freezing gates _ that stop changing their state is large .
in contrast if @xmath4 , the networks do not exhibit this kind of ordered behaviour ( see @xcite ) . the first analytical proof for this _
phase transition _ was given by derrida and pomeau ( see @xcite ) by studying the evolution of the hamming distance of random chosen initial states by means of so called _
annealed approximation_. the first proof for the number of freezing and ineffective gates was given by james lynch ( see @xcite , although slightly weaker results appeared earlier @xcite )
. depending on a parameter @xmath1 , that depends on the probabilities of the boolean functions , he showed that if @xmath5 almost all gates are ineffective and freezing , otherwise not .
although his analysis is very general , until now it was only applied to networks with connectivity @xmath6 and non - uniform probabilities for the boolean function : if the probability of choosing a constant function is larger or equal the probability of choosing a non - constant non - canalizing function ( namely the xor- or the inverted xor - function ) , @xmath1 is less or equal to one .
but it turns out that in some cases @xmath1 is equal to the expectation of the average sensitivity .
therefore we will first study the average sensitivity in section [ sec:02 ] .
afterwards it will be shown in section [ sec:03 ] how to use the results from the previous section to apply lynch s analysis to classical @xmath2-networks and _ biased _ random boolean networks .
but first we will give some basic definition used throughout the paper in section [ sec:01 ] .
in the following @xmath7 denotes the galois field of two elements , where addition , denoted by @xmath8 , is defined modulo 2 .
the set of vectors of length @xmath0 over @xmath9 will be denoted by @xmath10 .
if @xmath11 is a vector from @xmath10 , its @xmath12th component will be denoted by @xmath13 . with @xmath14
we will denote the _ unit vector _ which has all components zero except component @xmath12 which is one .
the hamming weight of @xmath15 is defined as @xmath16 and the hamming distance of @xmath17 as @xmath18 a boolean function is a mapping @xmath19 .
a function @xmath20 may be represented by its _ truth table _ @xmath21 , that is , a vector in @xmath22 , where each component of the truth table gives the value of @xmath20 for one of the @xmath23 possible arguments . to fix an order on the components of the truth table , suppose that its @xmath12th component equals the value of the corresponding function , given the binary representation ( to @xmath0 bits ) of @xmath12 as an argument .
in this section we will focus on the _ average sensitivity_. the average sensitivity is a known complexity measure for boolean functions , see for example @xcite .
it was already used to study boolean and random boolean networks for example in @xcite .
let @xmath20 denote a boolean function @xmath24 and @xmath25 a unit vector . 1 .
the sensitivity @xmath26 is defined as : @xmath27 2 .
the average sensitivity @xmath28 is defined as the average of @xmath26 over all @xmath29 : @xmath30 now consider the random variable @xmath31 , where @xmath32 denotes the set a all @xmath33 boolean function with @xmath0 arguments .
the probability measure is given by @xmath34 .
the expected value of the average sensitivity of this random variable is denoted by @xmath35 , and is given by @xmath36 the expected value was already derived in @xcite , and is given by : + let the random variable @xmath37 be defined as above , then @xmath38 [ theo:01 ] we will now concentrate on biased boolean functions .
the bias of a boolean function @xmath19 is defined as the number of @xmath39 in the functions truth table divided by @xmath40 . to define the bias of a random boolean function two definitions are possible .
first we can assumes that the truth tables of the boolean functions are produced by independent bernoulli trials with probability @xmath41 for a one ( this should be called _
mean _ bias , used for example in @xcite ) .
therefore consider the random variable @xmath42 .
the probability of choosing a function @xmath20 is given by @xmath43 for @xmath44 this is equivalent to the definition of @xmath37 . as a second possibility , we can only choose functions which have bias @xmath41 whereas to all other functions we assign probability 0 ( we will call this _ fixed _
bias ) . therefore consider the random variables @xmath45 .
denote the truth table of a function @xmath20 by @xmath21 .
further denote the set of all boolean functions @xmath20 with @xmath0 arguments and @xmath46 with @xmath47 .
the probability for a certain function chosen according @xmath48 is given by @xmath49 both definitions ensure that the expectation to get a one is equal to @xmath41 if the input of a function is chosen at random ( with respect to uniform distribution ) . but it will turn out that these two different methods of creating biased boolean functions , have a major impact on the average sensitivity .
the expectation of the average sensitivity of @xmath42 was derived in @xcite : let the random variable @xmath50 be defined as above : @xmath51 [ theo : expectedsensitivityb ] for the random variable @xmath48 we will now proof the following theorem : let the random variable @xmath48 be defined as above : @xmath52 [ theo : expectedsensitivity ] [ cor:01 ] to find @xmath53 we will first consider the random variable @xmath54 where @xmath55 and the probability of a function is given by @xmath56 consider the boolean functions as functions into @xmath57 by identifying @xmath58 with @xmath59 .
then we get or the function @xmath20 : @xmath60 where @xmath25 again denotes the unit vector with @xmath12th component set to @xmath39 .
hence by the linearity of the expectation @xmath61 now we form a matrix with the truth tables of all functions with hamming weight @xmath62 as column vectors : @xmath63 @xmath64 has exactly @xmath65 columns and @xmath23 rows .
each entry @xmath66 in the @xmath12th row and @xmath67th column equals the value of function @xmath68 given the binary representation of @xmath12 as input .
hence @xmath69 is determined by the number of @xmath39 in the row associated with @xmath70 divided by the length of the row .
consider an arbitrary row @xmath12 .
this row has a one at position @xmath67 if the corresponding column @xmath71 has a one at position @xmath12 .
but there are @xmath72 column vectors with a @xmath39 at position @xmath12
. it follows : @xmath73 as this holds for all @xmath70 , we have @xmath74 to find an expression for @xmath75 we consider two arbitrary rows @xmath76 ( @xmath77 ) .
define the following sum : @xmath78 obviously @xmath79 only if we have a @xmath39 in both rows at position @xmath12 .
this means for the column vectors @xmath80 of @xmath64 , we have @xmath81 .
but there are exactly @xmath82 such column vectors in @xmath64 .
therefore we have @xmath83 as @xmath84 for all @xmath85 it follows : @xmath86 hence substituting equations , and into equation leads to @xmath87 finally the claimed expression for @xmath53 can be obtained from the above equation by a substitution of @xmath62 : @xmath88 .
[ comment ] it should be noted , that the theorems [ theo:01 ] and [ theo : expectedsensitivityb ] can be proved using in a similar way .
also worth noting is the fact , that if the functions are chosen according @xmath89 or @xmath42 the expectation of the sensitivity of a fixed vector @xmath70 ( namely the expectation of @xmath90 ) is independent of @xmath70 ( see equation , , and ) .
hence the following lemma holds if @xmath91 or @xmath42 , then @xmath92 [ lemma:01 ] before proceeding to the next section , it should be noted , that using the same arguments as in the proof of theorem [ theo : expectedsensitivity ] , we can also prove the expectation of average sensitivity of order @xmath93 , defined as @xmath94 in this case , instead of summing up all unit vectors in equation , we sum up all vectors of hamming weight @xmath93 . as the equations and hold for all @xmath29 we conclude that @xmath95 and by similar arguments @xmath96 respectively @xmath97
as already mentioned james lynch gave a very general analysis of randomly constructed boolean networks ( see @xcite ) . before stating his results we give a formal definition for boolean networks a boolean network * b * is a 4-tuple @xmath98 where @xmath99 is a set of natural numbers , @xmath100 is a set of labeled edges on @xmath101 , @xmath102 is a ordered set of boolean functions such that for each @xmath103 the number of arguments of @xmath104 is the _ in - degree _ of @xmath105 in @xmath100 , these edges are labeled with @xmath106 , and @xmath107 .
suppose that a vertex @xmath12 has @xmath108 in - edges from vertices @xmath109 .
for @xmath110 we define @xmath111 the state of * b * at time 0 is called the _ initial state _ @xmath11 , so we define @xmath112 . for time
@xmath113 the state is inductively defined as @xmath114 .
hence we can in interpret @xmath101 as set of gates , @xmath100 and @xmath115 describes their functional dependence and @xmath11 is the networks initial state .
assume some ordering @xmath116 on the set of all boolean functions @xmath117 , where each function @xmath118 depends on @xmath108 arguments .
further a random variable @xmath119 with probabilities @xmath120 such that @xmath121 and @xmath122 .
now a random boolean network consisting of @xmath123 _ gates _ is constructed as follows : for each gate a boolean function is chosen independently , where the probability of choosing @xmath118 is given by @xmath124 .
suppose a function @xmath20 was chosen that has @xmath0 arguments , these arguments are chosen at random from all @xmath125 equally likely possibilities . at last
an initial state is chosen at random from the set on all equally likely states .
if the boolean functions are chosen according to our previously defined random variable @xmath37 we will call this networks @xmath2-networks with connectivity @xmath0 . if the functions are chosen according to @xmath48 or @xmath42 we will call this networks _ biased random boolean networks _ with connectivity @xmath0 and fixed bias @xmath41 respectively mean bias @xmath41 .
let us now state lynch s results .
his analysis depends on a parameter @xmath126 depending only on the functions and their probabilities .
we will define @xmath1 later in definition [ def : lambda ] .
first we have to state lynch s definition of _ freezing _ and _ ineffective _ gates : + let @xmath127 and @xmath103 . 1 .
gate @xmath105 freezes to @xmath110 in @xmath62 steps on input @xmath11 if @xmath128 for all @xmath129 .
2 . let @xmath130 .
+ a gate @xmath105 is @xmath62-ineffective at input @xmath15 if @xmath131 .
now we will state the main result .
+ let @xmath132 , @xmath133 be positive constants satisfying @xmath134 and @xmath135 where @xmath136 . 1 .
there is a constant @xmath137 such that for all @xmath127 @xmath138 when @xmath5 , @xmath139 and when @xmath140 , @xmath141 .
there is a constant @xmath137 such that for all @xmath127 @xmath142 when @xmath5 , @xmath139 and when @xmath140 , @xmath141 .
the above theorem shows that if @xmath5 almost all gates are freezing and ineffective and otherwise not .
the next corollary gives us more information what happens if @xmath140 : let @xmath140 . for almost all random boolean networks 1 .
if gate @xmath105 is not @xmath143-ineffective , there is a positive constant @xmath144 such that for @xmath145 , the number of gates affected by @xmath105 at time @xmath62 is asymptotic to @xmath146 , 2 .
if gate @xmath105 is not freezing in @xmath143 steps , there is a positive constant @xmath144 such that for @xmath145 , the number of gates that affect @xmath105 at time @xmath62 is asymptotic to @xmath146 .
now we will state the definition of @xmath1 for boolean networks : [ def : lambda ] let @xmath20 be a boolean function of @xmath0 arguments . for @xmath147 , we say that argument @xmath12 directly affects @xmath20 on input @xmath29 if @xmath148 .
now put @xmath149 as the number of @xmath12 s that directly affect @xmath20 on input @xmath70 . given a constant @xmath150 $ ] , we define @xmath151 obviously @xmath149 is identical to @xmath90 which will be used instead in the further discussion .
the constant @xmath152 is the probability that a random gate is one ( at infinite time ) given that all gates at time @xmath153 have probability @xmath154 of being one .
( see ( * ? ? ? * definiton 2 ) ) .
assume that we choose the functions according a random variable @xmath155 which should be either @xmath37 , @xmath48 or @xmath42 .
the functions are chosen out the set @xmath156 , we denote a function s probability with @xmath157 .
it follows that @xmath158 @xmath159 denotes the expectation of the sensitivity for a fixed @xmath70 , equation follows from lemma [ lemma:01 ] . therefore , together with theorem [ theo:01 ] and theorem [ cor:01 ] we proved the following : for random boolean networks , if 1 .
the functions are chosen according random variable @xmath42 , it follows that @xmath160 2 .
the functions are chosen according random variable @xmath48 , it follows that @xmath161 [ theo:02 ] as a special case of the above theorem we get ( or by using theorem [ theo:01 ] ) in random boolean networks , where the functions are chosen according to the random variable @xmath162 @xmath163
the results about @xmath2-networks are consistent with experimental results .
in fact if @xmath164 almost all networks almost all gates are freezing and almost all gates are ineffective and otherwise not ( see @xcite ) .
obviously , the border between the ordered and disordered phase is given by @xmath165 .
the resulting phase diagram for biased random boolean networks , where the functions are chosen according to @xmath48 and @xmath42 is shown in figure [ fig:01 ] .
it it interesting to note that if the functions are chosen with fixed bias , then also boolean networks with connectivity @xmath166 can become unstable . this conclusion can be drawn from lynch s original result already .
as mentioned in the introduction , he showed for @xmath166 , that @xmath140 if the probability of choosing a non - constant non - canalizing function , namely the xor or the inverted xor function , is larger than the probability of choosing a constant function .
for example if the bias is @xmath154 , the probability of choosing a constant function is zero , whereas both xor and inverted xor function have probability greater zero , hence @xmath140 .
( dashed ) and @xmath48 ( solid),title="fig : " ] it is interesting to compare our results with previous results obtained first by derrida and pomeau using the so called _ annealed approximation _ ( see @xcite ) . in their _ annealed model _ the functions and connections
are chosen at random at each time step .
considering two instances of the same annealed network starting in two randomly chosen initial states @xmath167 they show that @xmath168 where @xmath169 if @xmath170 and @xmath171 otherwise .
it is remarkable that the two models behave similar , but it is unclear whether this holds in general .
we would like to thank our colleges georg schmidt and stephan stiglmayr for proofreading and uwe schoening for useful hints . | in this work we consider random boolean networks that provide a general model for _ genetic regulatory networks_. we extend the analysis of james lynch who was able to proof kauffman s conjecture that in the _ ordered phase _ of random networks , the number of _ ineffective _ and _ freezing _ gates is large , where as in the _ disordered phase _ their number is small .
lynch proved the conjecture only for networks with connectivity two and non - uniform probabilities for the boolean functions .
we show how to apply the proof to networks with arbitrary connectivity @xmath0 and to random networks with _ biased _ boolean functions .
it turns out that in these cases lynch s parameter @xmath1 is equivalent to the expectation of _ average sensitivity _ of the boolean functions used to construct the network .
hence we can apply a known theorem for the expectation of the average sensitivity . in order to prove the results for networks with biased functions , we deduct the expectation of the average sensitivity when only functions with specific connectivity and specific bias are chosen at random .
random boolean networks , phase transition , average sensitivity pacs numbers : 02.10.eb , 05.45.+b , 87.10.+e |
odd - frequency ( of ) superconductivity @xcite , which is characterized by pair potential or pair amplitude with odd functions in time and frequency , has been recognized as a candidate of intriguing quantum states of matter .
while its realization has been theoretically proposed in a variety of systems @xcite there has been a long - standing problem : a thermodynamic instability arises if we apply a conventional approach to of superconductivity which has succeeded in describing ordinary even - frequency ( ef ) superconductors @xcite . at the same time , the sign of electromagnetic response function is reversed from the usual diamagnetic one , indicating a paramagnetic meissner response and negative superfluid density . therefore , its realization in bulk of condensed matter has been questioned . on the other hand , without such difficulties
the of superconductivity can exist as a surface state . while the ordinary ef
superconductivity is dominant in bulk , the of pairing state is present as an induced state @xcite .
it has been reported that the paramagnetic meissner response @xcite is observed in this induced of pairing state @xcite .
recent theoretical studies show that the of pairing can also be stabilized in a bulk , if we reconsider some conditions which are usually assumed in the theory of conventional superconductors .
namely , the sign arising from the of pair potential @xmath0 , which causes the thermodynamic instability , can be canceled by introducing an additional sign .
one of the solutions is to re - examine the conjugate relations of the frequency - dependent pair potential @xcite . using a path - integral formalism
, it has been shown in refs
. and that the problem can be resolved by using an unusual conjugate relation for pair potential , and consequently a description based on hermitian mean - field hamiltonian is impossible . subdominant even - frequency pairings induced in inhomogeneous systems such as surface or defect have been classified for this kind of of pairing based on symmetry arguments @xcite .
with this situation , however , recently josephson junctions have also been studied by fominov _
et al . _ and peculiar properties
are revealed @xcite . whereas a real current is obtained for the junction with ef superconductor ,
the current becomes imaginary if we make a junction with paramagnetic of superconductivity realized at e.g. the edges of @xmath1- or @xmath2-wave superconductors .
furthermore , an electromagnetic response function shows that the superfluid weight becomes complex number when these diamagnetic and paramagnetic odd - frequency pairings coexist @xcite .
these unphysical behaviors at least indicate that the conventional approach fails in describing the coexistence of the above of superconductor .
on the other hand , it has also been established that there is another type of stable of superconductivity , which can be described in the mean - field theory with hermitian hamiltonian . here , the additional minus sign to resolve the thermodynamic problem comes from a spatially oscillating phase of the pair amplitude @xcite , which is called staggered pairing @xcite .
the existence of staggered of pairing has been clearly demonstrated in the two - channel kondo lattice ( tckl ) @xcite . in this paper , for a deeper understanding of this type of of superconductors , we take tckl as a concrete example and explore novel properties of the junction systems to clarify the difference from already known superconductors . .
classification of uniform and staggered pairs .
the sign represents even or odd character of the exchange symmetry .
the symmetry in bracket is a secondarily induced pair from a purely of or ef pair potential . [
cols="^,^,^,^,^,^",options="header " , ] for transformed hamiltonian given in eq .
, the of pair has a uniform character . in terms of the classification in tab .
[ tab : list ] , the transformed state belongs to uniform oseo .
since the spin symmetry is broken in this picture , the uniform pair with eteo is mixed at the same time .
thus the secondarily induced pairs are transformed from esoo to eteo by the local gauge transformation . in the following
we explain how
the of pairs in tckl give the ordinary diamagnetic meissner effect , although the odd - frequency superconductors have long been considered to give a paramagnetic meissner kernel . while the numerical calculation @xcite shows the diamagnetic response , here we discuss it by focusing on the structure of the meissner kernel and do not enter the details . following the derivation in ref .
, only the anomalous part contributes to the meissner kernel @xmath3 which can be written in the form @xmath4 where @xmath5 is an electric charge , and we define the velocity @xmath6 along the @xmath7-direction .
we have also introduced the ` daggered ' anomalous green function by @xmath8 . from hermiticity of the hamiltonian
, we have the relation @xmath9 this relation can be explicitly shown by using the spectral representation .
we assume the inversion symmetry in the original lattice : @xmath10 .
let us consider the conventional spin - singlet @xmath11-wave ( ef ) superconductor as a reference .
the anomalous green s function has the structure in the form @xmath12 here we do not have to know the detailed functional form of @xmath13 . the orbital degree of freedom is not included here .
the meissner kernel is then given by @xmath14 for the velocity we have the relation @xmath15 , which gives the minus sign .
in addition , another sign comes from the spin factor @xmath16 , and hence in total the electromagnetic response is diamagnetic : @xmath17 . on the other hand ,
if we had @xmath11-wave spin - triplet of superconductivity , there would be no sign from spin - factor .
hence in this case the sign of the meissner kernel is reversed to give a paramagnetic response ( or sometimes called negative meissner effect ) . for @xmath1-wave superconductors ,
the minus sign comes from spatial part , i.e. @xmath18 , instead of spin part .
now we consider the kernel in tckl .
substituting eq . into eq . , we obtain @xmath19 although the factors from spin and orbital parts give the minus sign as @xmath20 , the sign operates twice and does not affect the total meissner kernel . for the velocity , we have @xmath21 originating from @xmath22 , which gives no minus sign in contrast to the above @xmath11-wave spin - singlet superconductor .
we further transform the expression in terms of ef and of pair amplitudes @xmath23 originating from eqs . and .
using the relation in eq . , the final expression is written as @xmath24 .
\label{eq : tckl_kernel}\end{aligned}\ ] ] namely the of pair ( oseo ) gives a diamagnetic contribution and the ef pair ( esoo ) shows a paramagnetic response , which is contrary to the standard wisdom . although it is not trivial to determine which parts give the dominant contribution , the numerical calculation shows that the of part is more dominant to give the total diamagnetic response @xcite .
this fact implies the importance of the of pair in tckl . the characteristic diamagnetic response by of pairs in tckl
is closely related to @xmath21 with finite center - of - mass momentum @xmath25 .
otherwise we would have another minus sign from @xmath26 and then the of pair gives paramagnetic contribution .
this point has also been numerically demonstrated in ref . .
at the end of this section , let us also consider the meissner kernel in the modified tckl given by eq . .
in this case the uniform pair amplitudes have the form @xmath27 .
\label{eq : pair_sym_mod}\end{aligned}\ ] ] the first and second terms in the right - hand side respectively correspond to oseo and eteo pairs .
the meissner kernel has the form @xmath28 instead of eq . .
the important point here is that the velocity is dependent on spin : @xmath29 and @xmath30 .
substituting eq . into the kernel
, we obtain the essentially same result as eq . which shows diamagnetic response .
while we have no staggered phase here , the additional minus sign comes from the spin - dependent velocity .
thus we have explicitly demonstrated that the staggered nature is not the only way to stabilize of superconductivity .
in this section we consider a tunneling conductance in normal metal ( n)/superconductor ( s ) junction , where s is a superconducting tckl in one dimension .
tunneling conductance can be calculated based on the blonder - thinkham - klapwijk theory @xcite , and a similar method is developed also in the tight - binding system @xcite .
we choose the bulk wave functions of tckl satisfying a proper boundary condition and calculate both the andreev and normal reflections in this n / s junction . for simplicity
we take the half - filled case ( @xmath31 ) in the following , and qualitatively same results can be obtained for @xmath32 . the mean - field hamiltonian introduced in the previous section
can be decomposed into two sets of subsystems : @xmath33 and @xmath34 with @xmath35 .
we focus on the former set , where the eigenenergies @xmath36 are given by @xmath37 and the corresponding eigenoperators by @xmath38 ( @xmath39 ) where @xmath40 with @xmath41 and @xmath42 .
the dispersion relation is illustrated in fig .
[ fig : dispersion ] .
when we take another subsystems , the behaviors discussed in this section remain unchanged .
we note that the gapless part @xmath43 contributes to the diamagnetic meissner kernel @xcite .
this is possible because the fermi surface is composed of both electron and hole to form a bogoliubov particle , as is distinct from an ordinary metal .
now we consider the n / s junction . the normal metal with @xmath44 and staggered of pairing state with @xmath45 in eq .
are placed at the left- ( @xmath46 ) and right - hand ( @xmath47 ) sides , respectively . when the ( @xmath48 ) conduction electron with the energy @xmath49 is injected from the left ,
the wave function in n is written in the vector form as @xmath50 the results for @xmath51 are obtained from the ones for @xmath49 by using the particle - hole symmetry . the coefficients @xmath52 and @xmath53 correspond to andreev and normal reflection weights , respectively . the wave vector @xmath54 is determined by the condition @xmath55 . here
we have only the two components because the localized pseudofermions are decoupled in n. a part of injected electron transmits into s , whose wave function is written as @xmath56 the wave vectors satisfy the relations @xmath57 and @xmath58 . here only the wave functions with positive group velocity appear .
we note that @xmath59 becomes imaginary for @xmath60 , where it exists as a quickly damping evanescent wave .
the n part at left and the s part at right are connected at the origin by the following tunnel hamiltonian : @xmath61 here we consider the barrier potential @xmath62 at the edge of the normal metal .
the present setup of the system is schematically illustrated in fig . [ fig:1d_junction](a ) . is considered at the edge of the normal metal in ( a ) . in
( b ) the josephson current is calculated at the two sites located at the center of the chain .
, width=302 ] we assume the sites for @xmath46 and for @xmath47 are described by the wave functions @xmath63 and @xmath64 , respectively . at the sites
@xmath65 , @xmath66 and @xmath67 we have the relations @xmath68 the diagonal matrices are made from eqs . and as @xmath69 , @xmath70 and @xmath71 .
these matrices are @xmath72 matrices , and they operate for the upper two components of @xmath73 , since pseudofermions have no inter - site hopping .
we also define the matrix @xmath74 which originates from the local part of eq . .
the function @xmath75 can not be described in general by either @xmath76 or @xmath73 due to the presence of the potential @xmath62 . to determine the coefficients we need to have another relations .
this situation is similar to a usual quantum mechanics which requires smooth wave functions at the boundary .
we consider the extrapolated wave functions @xmath77 and @xmath78 which satisfy the relations @xcite @xmath79 by solving these equations we can explicitly derive the coefficients @xmath52 , @xmath53 , @xmath80 , @xmath2 and the wave function @xmath75 at the interface .
the normal reflectance @xmath81 , andreev reflectance @xmath82 , and transmittances @xmath83 of quasiparticles are defined by @xmath84 which satisfy the sum rule of probability flow : @xmath85 note that the evanescent wave does not contribute to this sum rule . from these quantities
we define the conductance by @xmath86 where the factor @xmath87 originates from spin and orbital degrees of freedom and @xmath88 is the planck constant .
the condition @xmath89 means the existence of an excess current due to andreev reflection , or cooper pair tunneling into the s part .
we note that in actual systems the energy is given by @xmath90 with electric charge @xmath5 and bias voltage @xmath91 . ) and transmittances ( @xmath83 ) as a function of energy .
the transfer integrals at the junction are chosen as ( a ) @xmath92 and ( b ) @xmath93 .
the barrier potential is not included in this figure .
, width=302 ] we first discuss the n / s junction for the @xmath94 case .
figure [ fig : probability ] shows the reflectances and transmittances defined by eqs .
( [ eq : prob1][eq : prob4 ] ) .
we take the hybridization strength @xmath95 , and the gap is then @xmath96 . in the high transmissivity limit with @xmath92 for @xmath97 shown in fig .
[ fig : probability](a ) , a half of the injected electron transmits into the tckl superconducting state ( @xmath98 ) . the other half
is reflected into the normal metal both as electron ( @xmath99 ) and hole ( @xmath100 ) .
this behavior is in contrast to the ordinary @xmath11-wave superconductor , where the perfect andreev reflection ( @xmath101 ) can be observed . for small @xmath102 case ,
the energy dependence is modified while the behavior at low energy remains nearly unchanged .
the presence of normal reflection in tckl is related to the form of the mean - field hamiltonian given by eq . .
namely , the gapped structure in spectrum has the characters of both hybridization ( normal ) gap and superconducting ( anomalous ) gap .
consequently both the normal and anomalous self energies are present as in eqs . and , which cause normal and andreev reflections simultaneously .
another characteristic behavior different from ordinary superconductors is that the transmittance into the superconducting tckl is finite even at zero energy .
this is due to the presence of the fermi surface as shown in fig .
[ fig : dispersion ] .
hence bound states e.g. in s / n / s junction or at vortex core are unlikely formed even in the clean limit at low temperatures . ) and ( b ) attractive ( @xmath103 ) barrier potentials .
we take @xmath92 . , width=321 ] now we consider the situation with finite barrier potential at the edge of the normal metal .
the conductances are shown in fig .
[ fig : conductivity2 ] for @xmath92 , where we normalize them by the normal conductivity @xmath104 .
it is characteristic that the peaked structure is observed for @xmath105 while there is no such behavior for @xmath103 .
the effect of the sign of the barrier potential is remarkable near the gap edge ( @xmath106 ) , but it is irrelevant in the low - energy limit . the solutions at low energies can be obtained in a simple form . in the limit with @xmath107 , we can use the relations @xmath108 , @xmath109 and @xmath110 .
we then explicitly derive the reflectances for @xmath111 as @xmath112 , \\
b(e ) & \sim \frac 1 4 \left [ 1 - \left ( \frac v t \right ) \frac{e}{\gamma_0 } \right].\end{aligned}\ ] ] the magnitude of the reflectance of andreev reflection is enhanced with increasing energy @xmath113 for @xmath105 and is diminished for @xmath103 , while the normal reflection shows the inverse behavior .
thus the results are sensitive to the setup at the boundaries .
by contrast , the conventional spin - singlet @xmath11-wave superconductor does not show such a sign - sensitive behavior for barrier potential , and there is no difference between repulsive and attractive potentials . on the other hand , for sufficiently large magnitude of potentials both with @xmath105 and @xmath103 , the line shape of
the resulting conductance becomes similar to that of the local density of states ( ldos ) at the edge as will be shown in fig .
[ fig : dos](a ) .
this nonzero value of @xmath114 at zero energy clearly characterizes the present superconducting state as distinct from ordinary superconductors .
the present conductance in tckl is also different from that in spin - singlet @xmath2-wave or spin - triplet @xmath1-wave superconductor junctions @xcite . in these junctions , surface andreev bound state ( sabs )
produces a zero bias conductance peak @xcite and the magnitude of odd - frequency pairing amplitude is significant at the surface @xcite . on the other hand , in the present tunneling spectroscopy of tckl , the presence of the odd - frequency pairing does not produce a clear zero bias conductance peak .
the staggered of pairing state of tckl is coupled to the other types of superconductors in josephson junctions .
we consider the simple spin - singlet @xmath11-wave superconductivity ( esee ) , whose hamiltonian is given by @xmath115 in one dimension . the conduction electron part is written as @xmath116 . as we mentioned in the introduction
, there is a paramagnetic of superconductivity which is induced only at the edge from ef superconductivity in bulk . here as one of such examples we take the spin - triplet @xmath117-wave superconductor ( etoe ) .
the hamiltonian is explicitly written as @xmath118 when we make an edge with this hamiltonian , the local ( @xmath11-wave ) electron pair is generated at the edge .
since the spin structure of this pair is a triplet or even with respect to the spin exchange , the time dependence of the pair must be odd ( otee ) .
this induced of pair at the surface which is closely connected to the andreev bound state @xcite .
the josephson junction can be constructed by connecting tckl with one of the above superconductors .
the hamiltonian of this system is given by @xmath119 each of which describes the semi - infinite left - hand system ( @xmath46 ) , semi - infinite right - hand system ( @xmath120 ) and the middle junction part , respectively .
we take the spin - singlet @xmath11- or spin - triplet @xmath117-wave superconductors as @xmath121 and the tckl as @xmath122 .
the hamiltonian for the junction part is given by @xmath123 the present setup of the system in one dimension is schematically illustrated in fig .
[ fig:1d_junction](b ) .
the josephson current @xmath124 is calculated at the center of this junction : @xmath125 here the josephson current is well defined because the gauge - symmetry breaking terms are not included at the junction region , and the equation of continuity locally holds only by quasiparticle flow .
the josephson current can be calculated by using the semi - infinite green function @xcite . as an alternative method one can approximate this by the green function at the edge of the finite chain .
we take the number of sites as @xmath126 in the following .
the semi - infinite left- and right - hand surface green functions @xmath127 and @xmath128 are explicitly derived from the hamiltonians and , which can be written in a nambu matrix form with respect to spin / orbital index .
the local green functions at the site @xmath66 and @xmath129 without the connection by @xmath130 are given by @xmath131^{-1 } , \label{eq : gl0 } \\
\hat g^{\rm r}_1 & = [ z\hat 1 - \hat { \gamma } ' \hat g^{\rm r}_{\infty } ( \hat { \gamma}')^{\dagger}]^{-1 } , \label{eq : gr1}\end{aligned}\ ] ] respectively . the indices 0 and 1 mean the site index at the junction part .
the matrices @xmath132 , @xmath133 and @xmath134 are made from eq . in a manner similar to n / s junction . using these quantities ,
the green functions at the junction are given by @xmath135^{-1 } , \label{eq : g10 } \\
\hat g_{01 } & = \hat g^{\rm l}_{0 } \hat { \gamma}_{\rm i } [ ( \hat g^{\rm r}_1)^{-1 } - \hat { \gamma}_{\rm i}^{\dagger}\hat g^{\rm l}_{0 } \hat { \gamma}_{\rm i}]^{-1 } .\label{eq : g01}\end{aligned}\ ] ] the josephson current defined in eq
. is then calculated at finite temperatures from @xmath136 .
-wave superconductor in ( c ) and ( d ) .
the parameters are chosen as @xmath137 for ( a , b ) and @xmath138 for ( c , d ) .
the infinitesimal imaginary part @xmath139 is taken as @xmath140 .
, width=283 ] first we show the ldos and local pairing amplitude at the edge of the semi - infinite chain calculated from @xmath141 and @xmath142 .
figures [ fig : dos](a ) and ( b ) show the ldos proportional to @xmath143 and pair amplitude @xmath144 , respectively , for tckl .
the values are normalized by @xmath145 which is the density of states for a normal metal .
in contrast to the conventional spin - singlet @xmath11-wave superconductor , the ldos is nonzero at the fermi level .
this is because the half of the bogoliubov particles in the of pairing state have an energy gap and the others still have the fermi surface as shown in fig .
[ fig : dispersion ] .
the frequency dependence of the real part of pair amplitude ( or anomalous green function ) shown in fig .
[ fig : dos](b ) is odd with respect to real frequency .
these behaviors are similar to the ones in bulk @xcite .
although here we can not see ef components , it appears as the inter - site green functions .
on the other hand , the ldos at the edge of the spin - triplet @xmath117-wave superconductor has the sharp peak as shown in fig .
[ fig : dos](c ) , which is known as a consequence of the andreev bound state @xcite .
this non - trivial localized edge state is formed when the sign of the gap function felt by quasiparticle is reversed at the reflection process .
figure [ fig : dos](d ) displays the local pair amplitude which is odd in frequency ( otee ) , although in bulk only the spin - triplet @xmath117-wave ef pair ( etoe ) is formed @xcite . .
the parameters are same as the ones in fig . [
fig : dos ] .
, width=283 ] let us discuss how the edge state is connected to the bulk state .
figure [ fig : spat_dos ] shows the spatial dependence of the ldos at the fermi level . here the ldos at the fermi level has the spatial dependence in the form @xmath146 where @xmath147 and @xmath148 are smooth functions in space .
this oscillating behavior in a staggered manner originates from half - filled situation with @xmath149 , and the period of oscillation changes for @xmath32 reflecting the change of fermi wave vectors . as shown in fig .
[ fig : spat_dos](a ) , the ldos at the edge of tckl is continuously connected to the bulk .
a slow relaxation is characteristic for the metallic state , and is consistent with the presence of fermi surface in superconducting state of tckl .
( in numerical simulation , the healing length , which may be physically regarded as mean free path , is given by @xmath150 with small but finite @xmath139 . )
hence the character of this zero - energy state can be regarded as similar to the one in bulk tckl . for the spin - triplet @xmath117-wave superconductor shown in fig .
[ fig : spat_dos](b ) , on the other hand , the zero - energy state located at the edge vanishes quickly as we go into the bulk state .
this edge state has a different character from the bulk state in this case . with these preliminaries ,
now we consider the josephson junction . in the following we consider the zero barrier potential case ( @xmath94 ) unless explicitly stated otherwise .
the phase of the pair amplitude in the left - hand system is taken as @xmath151 , while it is set as zero in the right - hand system as illustrated in fig .
[ fig:1d_junction ] .
we begin with the spin - singlet @xmath11-wave superconductor / tckl junction .
however , the josephson current completely vanishes in the present simple setup .
as explained later , the absence of josephson current is related to the fact that symmetries of the induced pairs located at the edges do not match between the left- and right - hand sides . in order to have finite current ,
the simplest modification without changing bulk properties , is to change the tunnel matrix at the interface as @xmath152 with @xmath153 .
we call this the setup ( i ) .
note that with this tunnel matrix both the spin and orbital symmetries are broken but their product is not broken ( see also tab .
[ tab : induced_pair ] ) . on the other hand ,
the more realistic setup giving finite currents is to modify the bulk nature of tckl with keeping the tunnel matrix @xmath154 .
we consider the orbital field both for conduction electrons and localized pseudospin , whose hamiltonain is given by @xmath155 ( called the setup ( ii ) in the following ) this term breaks the orbital symmetry and experimentally corresponds to the uniaxial pressure effect .
when we make the junction in real materials , some stress should be applied to the edge of tckl .
hence the effect of eq .
will reasonably appear .
for simplicity we take @xmath156 in the following , but this assumption does not affect qualitative aspect of the results .
dependences of the josephson current @xmath124 for ( a ) spin - singlet @xmath11-wave and ( b ) spin - triplet @xmath117-wave superconductors connected to tckl with the setup ( i ) [ @xmath157 , @xmath158 .
the results for the setup ( ii ) [ @xmath159 , @xmath160 are shown in ( c ) and ( d ) .
the parameters are chosen as @xmath138 , @xmath161 , @xmath162 and @xmath163 .
the results for finite barrier potentials ( @xmath164 ) are also shown .
, width=321 ] figure [ fig:1d_current ] shows the phase @xmath165 dependence of josephson currents , which is normalized by @xmath166 .
let us first discuss the case with the setup ( i ) . as shown in fig .
[ fig:1d_current](a ) , the josephson current has the functional form of @xmath167 for spin - singlet @xmath11-wave superconductor / tckl junction .
this indicates that the first - order coupling vanishes in this case .
the josephson current for the spin - triplet @xmath117-wave superconductor / tckl junction have the form @xmath168 as seen in fig . [ fig:1d_current](b ) .
on the other hand , the results are changed for the setup ( ii ) as shown in fig .
[ fig:1d_current](c , d ) .
the first - order coupling survives for @xmath11-wave superconductor / tckl junction , while it vanishes in the junction using @xmath117-wave superconductor .
these behaviors can be qualitatively understood by considering the two - site model ( zero - dimensional system ) that simulates the edges of right- and left- superconductors .
here we focus on the case with the setup ( ii ) , and the more detailed analysis including the setup ( i ) is given in appendix .
the local spin - singlet @xmath11-wave pairing field term is given by eq .
, and the pairing field for tckl by eq . .
we directly connect these two sites by the matrix defined by the third line of eq . .
the josephson current is explicitly evaluated as @xmath169 within the lowest - order approximation . the left- and right - anomalous green functions
are given by @xmath170 where @xmath171 is the orbital - dependent hybridization function .
for @xmath172 the anomalous green function of tckl is a purely odd function with respect to frequency , but the even - frequency component mixes in the presence of orbital fields . from these expressions
it is clear that the josephson current becomes zero if we take zero orbital field ( @xmath172 ) . with finite orbital field , on the other hand , @xmath173 and @xmath174 have the same parity in frequency space , and the first - order contribution to
the josephson current becomes finite .
namely , the induced ef pair in tckl contributes to the josephson coupling . for spin - triplet @xmath117-wave superconductor / tckl junction ,
the josephson coupling is expressed by odd - frequency spin - triplet @xmath11-wave ( otee ) and odd - frequency spin - singlet @xmath11-wave ( oseo ) pairing .
then , the first - order contribution with respect to @xmath175 vanishes .
thus we obtain consistent results with numerical calculations for a chain discussed above .
next we discuss the above josephson junction from symmetry point of view .
originally , the oseo+esoo pairs are present in tckl without any field as discussed in sec . iii .
on the other hand , for spin - singlet @xmath11-wave superconductor the esee pair and the induced osoe pair are present at the edge . in a similar manner
the etoe and induced otee pairs exist for spin - triplet @xmath117-wave superconductor .
thus , no symmetries match between tckl and the other superconductors , and the first - order coupling in josephson junction becomes zero .
in fact , this vanishing current persists to higher orders .
to explain this behavior , we must specify the component of orbital - triplets in @xmath11- and @xmath117-wave superconductors .
in the present setup , since we do not include the orbital degrees of freedom explicitly , the triplet component is identified as @xmath176 or @xmath177 and no @xmath178 component .
thus the mismatch between orbital - singlet in tckl and orbital - triplet ( @xmath179 ) in @xmath11- and @xmath117-wave superconductors gives exactly zero current in the present system . with the tunnel matrix in the setup ( i ) ,
the spin - orbital symmetry breaking is present and the induced pair is otee+etoe according to tab .
[ tab : induced_pair ] .
hence , the first - order josephson coupling survives for @xmath117-wave superconductor / tckl junction , but it vanishes for @xmath11-wave superconductor / tckl case . similarly , with uniaxial pressure in the setup ( ii ) , the esee+osoe pairs are newly generated at the edge of tckl , where orbital - triplet component include @xmath176 .
hence the first - order contribution to josephson current becomes nonzero for tckl / spin - singlet @xmath11-wave superconductor junction .
since we rely only on the symmetry of cooper pairs , the above discussion should be applicable also to systems in higher dimensions .
finally we make a comment on the effect of a barrier potential @xmath62 at the junction part ( @xmath180 ) .
the phase @xmath165 dependence of the currents with repulsive and attractive potentials are shown in fig .
[ fig:1d_current ] .
the functional forms are not influenced qualitatively by the sign of @xmath62 , since the barrier potential does not create any new species of cooper pairs .
in addition , we do not observe the difference between @xmath105 and @xmath103 .
this behavior is consistent with results in the n / s junction : the sign of the barrier potentials does not affect the behaviors in the low - energy limit as shown in fig .
[ fig : conductivity2 ] .
we have investigated the staggered of pairing realized in tckl from a symmetry point of view . although the pair potential is purely odd function with respect to time ( frequency ) , both of and ef components of pair amplitude are present due to the absence of translational invariance even in the bulk .
the existing pairs in bulk are identified as primary oseo and secondary esoo .
we have also shown that a local gauge transformation changes the staggered state into uniform one with spin - symmetry broken state .
the mechanism for the diamagnetic meissner effect has been explained by focusing on the symmetry of pair amplitude and structure of the meissner kernel .
in addition to time / spin / space / orbital structures of cooper pairs , the finite center - of - mass momentum , which affects the sign of the velocity , plays an important role for diamagnetic response .
the n / s junction has been constructed and it is shown that the normal reflection is always present in addition to andreev reflection .
this behavior is in contrast with ordinary bcs superconductors , where only the andreev reflection is observed for high transmissivity limit .
the difference lies in the presence of normal self energy in the superconducting state of tckl . due to a finite density of states ,
the transmittance into tckl is also nonzero even at low energies .
hence the bound state at e.g. superconducting vortex core is unlikely to be formed .
when we consider the barrier potential at the interface , the conductance shows the difference between attractive and repulsive potentials , although no such difference is observed in conventional superconductors .
we have also investigated the josephson junction using green function formalism .
we connect tckl both with spin - singlet @xmath11-wave and spin - triplet @xmath117-wave pairing states . here a uniaxial pressure effect is considered for tckl , which is naturally expected at the edge of real materials . for tckl / spin - singlet @xmath11-wave superconductor junction , the relative phase @xmath165 dependence of josephson current becomes @xmath181 .
it becomes @xmath182 for tckl/@xmath117-wave superconductor junction , where no first - order coupling appears .
these josephson currents can be qualitatively described by a zero - dimensional system .
the symmetry of the pairs induced at the edge is a key ingredient to understand the current phase relations of josephson junctions .
this work was financially supported by a grant - in - aid for jsps fellows ( grant no .
13j07701 ) , a grant - in aid for scientific research on innovative areas `` topological material science '' ( grant no .
15h05853 ) , and a grant - in - aid for scientific research b ( grant no .
15h03686 ) .
we consider the simple two - site model given by @xmath183 where @xmath184 here @xmath121 and @xmath122 simulate the edge of the @xmath11-wave superconductor and tckl , respectively . the tunnel matrix is given by eq . , and the current is simply defined by @xmath185 we can solve this model analytically , which helps us understand the basic properties of josephson junction .
we define the green function @xmath186 the first component of this vector is relevant to current .
its fourier transformation @xmath187 satisfies the equation @xmath188 where @xmath189 is the constant vector and @xmath190 the other contributions to current can also be calculated in a similar manner . from the above equations ,
we can obtain eq . .
for the special case with @xmath191 and @xmath192 , namely without any symmetry breaking fields , the green function matrix @xmath193 becomes diagonal .
correspondingly , the anomalous parts in the second term of the eq . , which is relevant to higher - order josephson couplings , are effectively dropped from the equation and the josephson coupling terms vanish in general .
this behavior is consistent with the results discussed in sec .
v. v.l .
berezinskii , jetp lett . * 20 * , 287 ( 1974 ) .
kirkpatrick and d. belitz : phys .
lett . * 66 * , 1533 ( 1991 ) .
d. belitz and t.r .
kirkpatrick : phys .
b * 46 * , 8393 ( 1992 ) . a. balatsky and e. abrahams , phys . rev .
b * 45 * , 13125 ( 1992 ) .
v.j . emery and s. kivelson , phys .
b * 46 * , 10812 ( 1992 ) .
m. vojta and e. dagotto : phys .
b * 59 * 713 ( 1999 )
. y. fuseya , h. kohno , and k. miyake : j. phys .
jpn . * 72 * 2914 ( 2003 ) .
t. hotta : j. phys .
. jpn . * 78 * 123710 ( 2009 ) .
k. shigeta , s. onari , k. yada , and y. tanaka : phys .
b * 79 * 174507 ( 2009 ) ; k. shigeta , y. tanaka , k. kuroki , s. onari , and h. aizawa : phys . rev .
b * 83 * 140509 ( 2011 ) .
j. otsuki : phys .
lett . * 115 * , 036404 ( 2015 ) .
p. coleman , e. miranda , and a. tsvelik , phys .
lett . * 70 * , 2960 ( 1993 ) .
r. heid , z. phys .
b * 99 * , 15 ( 1995 ) .
bergeret , a.f .
volkov , and k.b .
efetov , rev .
phys . * 77 * , 1321 ( 2005 ) .
y. tanaka and a.a .
golubov , phys .
. lett . * 98 * , 037003 ( 2007 ) .
s. higashitani , s. matsuo , y. nagato , and k. nagai , phys .
b * 85 * , 024524 ( 2012 ) .
y. tanaka , m. sato , and n. nagaosa , j. phys .
. jpn . * 81 * , 011013 ( 2012 ) . a. di bernardo , z. salman , x. l. wang , m. amado , m. egilmez , m. g. flokstra , a. suter , s. l. lee , j. h. zhao , t. prokscha , e. morenzoni , m. g. blamire , j. linder and j. w. a. robinson .
x , * 5 * , 041021 , ( 2015 ) .
d. belitz and t.r .
kirkpatrick , phys .
b * 60 * , 3485 ( 1999 ) .
d. solenov , i. martin , and d. mozyrsky , phys . rev .
b * 79 * , 142502 ( 2009 ) .
h. kusunose , y. fuseya , and k. miyake , * 80 * , 054702 ( 2011 ) .
y. asano , y.v .
fominov , and y. tanaka , phys .
b * 90 * , 094512 ( 2014 ) .
fominov , y. tanaka , y. asano , and m. eschrig , phys .
b * 91 * , 144514 ( 2015 ) .
v. martisovits and d.l .
cox , phys .
b * 57 * , 7466 ( 1998 ) .
r. heid , ya.b .
bazaliy , v. martisovits , and d.l .
cox , phys .
* 74 * , 2571 ( 1995 ) .
s. hoshino and y. kuramoto , phys .
lett . * 112 * , 167204 ( 2014 ) . s. hoshino , phys .
b * 90 * , 115154 ( 2014 ) .
a.m. black - schaffer and a.v .
balatsky , phys .
b * 88 * , 104514 ( 2013 ) .
y. asano and a. sasaki , phys .
b * 92 * , 224508 ( 2015 ) . c. triola and a.v .
balatsky , arxiv:1512.07031 ( 2015 ) .
d. podolsky , e. demler , k. damle , and b. i. halperin , phys .
b * 67 * , 094514 ( 2003 ) .
chen , o. vafek , a. yazdani , and s .- c .
zhang , phys .
lett . * 93 * , 187002 ( 2004 ) .
f.j . ohkawa , phys .
b * 73 * , 092506 ( 2006 ) .
k. seo , h .-
chen , and j. hu , phys .
b * 78 * , 094510 ( 2008 ) .
d.f . agterberg and h. tsunetsugu , nat .
phys . * 4 * , 639 ( 2008 ) .
e. berg , e. fradkin , and s.a .
kivelson , phys .
lett . * 105 * , 146403 ( 2010 ) .
t. yoshida , m. sigrist , and y. yanase , phys . rev .
b * 86 * , 134514 ( 2012 ) .
cho , r. soto - garrido , and e. fradkin , phys .
lett . * 113 * , 256405 ( 2014 ) .
yang , phys .
lett . * 63 * , 2144 ( 1989 ) .
r.r.p . singh and r.t .
scalettar , phys .
* 66 * , 3203 ( 1991 ) .
j. de boer , v.e .
korepin , and a. schadschneider , phys .
lett . * 74 * , 789 ( 1995 ) .
japaridze , a.p .
kampf , m. sekania , p. kakashvili , and ph .
brune , phys .
b * 65 * , 014518 ( 2001 ) .
h. zhai , phys .
b * 71 * , 012512 ( 2005 ) .
m. jarrell , h. pang , d.l .
cox , and k.h .
luk , phys .
lett . * 77 * , 1612 ( 1996 ) .
cox and a. zawadowski : adv .
phys . * 47 * , 599 ( 1998 ) .
p. fulde , and r.a .
ferrell , phys .
rev . * 135 * , a550 ( 1964 ) .
a.i . larkin , and y.n .
ovchinnikov , sov .
jetp * 20 * , 762 ( 1965 ) .
blonder , m. tinkham , and t.m .
klapwijk , phys .
b * 25 * , 4515 ( 1982 ) .
zhu and h. kroemer , phys .
b * 27 * , 3519 ( 1983 ) .
a.v . burmistrova , i.a .
devyatov , a.a .
golubov , k. yada , and y. tanaka , j. phys .
. jpn . * 82 * , 034716 ( 2013 ) .
s. kashiwaya and y. tanaka , rep .
* 64 * 1641 ( 2000 ) . c. r. hu , phys .
* 72 * , 1526 ( 1994 ) .
y. tanaka and s. kashiwaya , phys .
lett . * 74 * 3451 ( 1995).y .
tanaka , a. a. golubov , s. kashiwaya , and m. ueda , phys.rev .
lett . * 99 * , 037005 ( 2007 ) . | junction systems of odd - frequency ( of ) superconductors are investigated based on a mean - field hamiltonian formalism . one - dimensional two - channel kondo lattice ( tckl ) is taken as a concrete example of of superconductors .
properties of normal and andreev reflections are examined in a normal metal / superconductor junction . unlike conventional superconductors ,
normal reflection is always present due to the normal self energy that necessarily appears in the present of pairing state .
the conductance reflects the difference between repulsive and attractive potentials located at the interface , which is in contrast with the preexisting superconducting junctions .
josephson junction is also constructed by connecting tckl with the other types of superconductors .
the results can be understood from symmetry of the induced cooper pairs at the edge in the presence of spin / orbital symmetry breaking .
it has also been demonstrated that the symmetry argument for cooper pairs is useful in explaining meissner response in bulk . |
as is well known , the greisen - zatsepin - kuzmin ( gzk ) cutoff constrains detected uhecrs to have been produced in or near the galaxy . specifically , detected proton primaries with energies exceeding @xmath0 ev
must have been produced within 50 mpc of earth , and nuclei propagation distances are constrained even further @xcite .
the near - isotropy of detected uhecr arrival directions , however , suggests that galactic source locations are not easily associated with the observed arrival directions .
we attempt to reconcile these observations by examining the possibility that uhecrs are galactic in origin , but consist of iron nuclei primaries with trajectories influenced by the galactic magnetic field .
it has been shown by @xcite that mhd winds from young neutron stars are capable of galactic production of iron primaries that fit the observed uhecr energy spectrum .
such a production mechanism implies that most source locations should be found within the galactic disk . with this in mind
, we propagate iron nuclei uhecrs through a realistic model of the galactic magnetic field to investigate the possibility that galactic sources and isotropic arrival directions are not mutually exclusive phenomena .
following @xcite and @xcite , we adopt a large - scale regular galactic magnetic field associated with the spiral arms of the galaxy . specifically , we choose a bisymmetric even - parity field model ( bss - s ) in which the field reverses direction between different spiral arms , but is symmetric with respect to the galactic plane . the field strength in the plane , directed along the spiral arms , at a point @xmath1 in galactocentric coordinates
is given by @xmath2 with @xmath3 kpc as the galactocentric distance to maximum field strength at @xmath4 with @xmath5 , where the pitch angle is @xmath6 .
the radial dependence of the field strength is given by @xmath7 where @xmath8 kpc is the galactocentric distance to the sun and @xmath9 kpc is a smoothing factor that allows for @xmath10 behavior beyond 4 kpc from the galactic center .
the field equation in the galactic plane is given by @xmath11\ ] ] for this even - parity model , we introduce a @xmath12 dependence of the following form @xmath13 where the values @xmath14 kpc and @xmath15 kpc reflect the scale heights of the field in the galactic disk and halo , respectively . figure ( 1 ) gives a graphical representation of this regular field model , clearly illustrating the field association with the spiral arms .
note that this model generates field values ranging approximately from @xmath16 g , with a value of @xmath17 g in the solar neighborhood .
we develop two distinct approaches to the simulation of galactic uhecr trajectories . in our first method , we model a distribution of galactic sources and emit uhecrs according to an assumed @xmath18 emission energy spectrum in order to study the arrival energy spectrum at a detector .
this allows for comparison between the injected and observed energy spectrum .
our second approach consists of modeling the antiparticle trajectories as they depart earth , given the detected uhecr arrival spectrum .
this is equivalent to plotting particle trajectories that are guaranteed to intersect earth , and this method has the advantage of allowing us to investigate potential source locations corresponding to real observed arrival directions .
our simulation of uhecr particle trajectories proceeds from a reasonable distribution of neutron stars within the galactic disk to serve as injection regions for uhecrs . for simplicity
, we assume that all sources are located in the galactic disk of radius 25 kpc and thickness .65 kpc ( taken from the scale height of the thin disk ) . we can derive a reasonable number of distinct sources from @xmath19 where @xmath20 myr is the program integration time and @xmath21 is the neutron star birth rate . these values produce @xmath22 distinct uhecr sources .
given this set of sources , we assign a random time @xmath23 at which each source emits randomly directed uhecrs according to the @xmath24 energy spectrum , as is expected from neutron star sources @xcite .
the particles then propagate through the galaxy with paths influenced by the lorentz force as they traverse the galactic magnetic field .
since an earth - sized detector is too small to detect a significant number of events with computationally reasonable numbers of injections , we develop a series of larger detectors .
the first is a 2d galactocentric cylindrical detector of radius @xmath8 kpc , spanning 20 - 40 pc above the galactic plane .
the second detector is a 2d galactocentric ring located 30 pc above the galactic plane with an inner radius of 8.49 kpc and an outer radius of 8.51 kpc .
these 2d detectors , placed at the solar distance from the galactic center , possess a number of advantages over local 3d detectors by generating a much higher detection flux while simultaneously reducing the bias for detection of local sources .
with these detectors defined , we sample uhecr energy spectra to compare the detected energy spectrum with the assigned @xmath24 injection energy spectrum . to accomplish this task , we inject millions of particles from our 300,000 sources and plot the detected energy spectrum .
figure ( 2 ) shows a typical detection energy spectrum from such a simulation .
since the energy emission spectrum is continuous , we bin the detected data before fitting it to a function of the form @xmath25 . for the data shown , we calculate @xmath26 .
thus , we find that the emission and arrival spectra are not significantly different for uhecrs .
this is an important result since it is unknown _ a priori _ if the detection energy spectrum should reflect the emission spectrum when particles have the potential to be trapped in the galactic field .
, matching the emission spectrum.,width=321,height=321 ] this simulation addresses the other end of the uhecr problem , propagating anti - iron nuclei from a distribution of arrival directions at earth . in this approach ,
we assign energies to the emitted antiparticles and propagate them through the galaxy to determine possible source locations for the corresponding uhecr particles .
figure ( 3 ) shows one such sample set of trajectories for the case of isotropically distributed 50 eev iron nuclei .
this method guarantees that our trajectories intersect earth , but we are now left to evaluate source location without an initial neutron star distribution . keeping in mind the need for galactic neutron stars as sources , we define a potential uhecr source location as any point along an antiparticle trajectory that intersects the galactic disk .
furthermore , we closely examine those antiparticle trajectories for which the path first leaves and then reenters the disk .
these particular paths are the least likely to be easily identified with their galactic sources , and their abundance would lend support to galactic uhecr origins .
we proceed by first emitting a known distribution of anti - iron nuclei from earth . specifically
, we choose to trace back agasa events with detected energies that exceed 100 eev .
then , using the agasa energies and arrival directions , we send the particles back through the galactic field .
since most of the detected high - energy agasa events are observed to have arrived away from the plane of the disk , we do nt have to worry about the exact difference between arrival direction and source location .
any path that manages to reintersect the disk some time after emission will have deviated enough to be of interest to us .
figure ( 4 ) shows the set of trajectories modeled from those agasa events above 100 eev .
it is clear that there is little turning back towards the galactic disk once the particles have exited .
the only path that remains in the galactic disk for an extended period of time corresponds to the agasa particle detected nearest the disk . with our choice of field parameters ,
the agasa events do not clearly point back to sources within the galactic disk .
galactic field.,width=321,height=321 ]
needless to say , the results of both types of simulation exhibit great sensitivity to certain parameters of the regular magnetic field model .
specifically , we find that variations in field strength strongly affect possible source locations for detected uhecrs .
figure ( 5 ) shows a contrast between the source location distributions for local field values of 3 and 6 @xmath27 .
the 3 @xmath27 field results in an abundance of detections involving local sources , while the 6 @xmath27 case exhibits pronounced non - local galactic features in the detection spectrum .
furthermore , differences in the field strength do greatly alter our agasa trajectories .
figure ( 6 ) shows the trajectories produced from agasa information with a 6 @xmath27 galactic field .
clearly , our choice of field parameters is important .
in addition to the regular galactic field used in our models , there are a number of additional magnetic field components that may be present in the galaxy .
in particular , a galactic wind similar in nature to the solar wind can significantly alter uhecr trajectories if the wind field strength is comparable to the regular field strength .
such a wind model with an azimuthal field strength of 7 @xmath28 has been proposed to redirect the highest energy particles such that a single source in virgo can be detected as an isotropic distribution on earth @xcite .
this strong wind effectively funnels isotropic arrival directions to the north galactic pole for a narrow range of energies . for iron nuclei
, this strong version of the galactic wind guarantees diffusive behavior up to the highest energies detected thus far .
local high - field regions within the galaxy also have the potential to contribute to the diffusion of uhecrs .
molecular clouds , for instance , are typically associated with mg fields that could serve as scattering regions for cosmic rays .
our preliminary research has indicated that the known distribution of molecular clouds has a small enough filling factor to cause little change in the paths of most uhecrs , but these and other magnetic field inhomogeneities must be better understood before the picture of uhecr propagation can be made complete . | we consider the effects of the galactic magnetic field on the propagation of ultra high energy cosmic rays ( uhecrs ) . by employing two methods of trajectory simulation
, we investigate the possibility that uhecrs are produced within the galaxy with paths strongly influenced by the galactic magnetic field .
such trajectories have the potential to reconcile the existing conflict between proposed local sources and isotropic uhecr arrival directions . |
let @xmath6 be hyperplanes in general position in complex projective space @xmath7 @xmath8 .
being in general position simply means that any @xmath9 hyperplanes have empty intersection . by a theorem of dufresnoy (
viii ) , there exists a constant @xmath10 depending only on the hyperplanes and the dimension @xmath11 such that if @xmath3 is a holomorphic map from the unit disc to @xmath1 which omits the hyperplanes @xmath12 then @xmath13 where @xmath5 denotes the norm of the derivative of @xmath3 with respect to the fubini - study metric on @xmath1 and can be defined by @xmath14 where @xmath15 for some choice of homogeneous coordinate functions for @xmath16 in modern terms ( see for instance @xcite or @xcite ) , dufresnoy s theorem says that the complement of @xmath0 hyperplanes in general position in @xmath1 is complete hyperbolic and hyperbolically embedded in @xmath17 in @xcite , dufresnoy remarks that the constant @xmath2 depends on the hyperplanes @xmath18 in a `` completely unknown '' way .
our purpose is to give an explicit estimate for the constant @xmath19 and therefore an explicit lower bound on the ratio of the infinitesimal kobayashi metric ( a. k. a. royden s function @xcite ) on the hyperplane complement as compared to the fubini - study metric on projective space . to that end , let @xmath20 $ ] be homogeneous coordinates on @xmath7 and let @xmath21 be @xmath9 hyperplanes in general position in @xmath1 given by linear defining forms @xmath22 normalized such that @xmath23 consider the @xmath24-matrix @xmath25 and let @xmath26 be the @xmath9 eigenvalues of the matrix @xmath27 listed in non - decreasing order .
we then define the following quantities : @xmath28 we can now state our main theorem as let @xmath29 be @xmath0 hyperplanes in general position in @xmath17 let @xmath30 and @xmath31 let @xmath32 be the binomial coefficient @xmath33 and @xmath34 let @xmath3 be a holomorphic map from the unit disc @xmath35 to @xmath1 omitting @xmath36 then , @xmath37 when @xmath38 the classical theorem of landau @xcite gives effective upper bounds , and very good bounds were obtained in @xcite and @xcite .
effective , though very non - sharp , estimates can also be derived from nevanlinna s theory with precise error terms @xcite .
for some symmetrical arrangements of the omitted points , one even knows sharp estimates @xcite , @xcite .
dufresnoy s proof of his theorem was based on the deep results of bloch and cartan @xcite , ( * ? ? ?
it is not clear whether this approach can give an effective estimate .
one attempt is this direction is @xcite .
the theorem of dufresnoy can also be obtained from borel s theorem @xcite by using a compactness argument called the zalcman - brody rescaling lemma @xcite , @xcite .
this proof does not give any effective estimate of @xmath39 if @xmath3 omits @xmath40 instead of @xmath41 hyperplanes , there is an estimate of cowen @xcite ( see also @xcite ) , which he gets by constructing negatively curved metrics on the hyperplane complement .
it seems difficult to adapt this method to the case of @xmath0 omitted hyperplanes , except of course for @xmath42 when they are the same . for arbitrary @xmath11
there is also a negative curvature approach of babets @xcite , @xcite , but babets s estimates involve higher order derivatives or higher associated curves . in this paper , we obtain an explicit estimate by putting in a quantitative form the argument from @xcite , where another proof of dufresnoy s theorem was given .
the proof in @xcite was based on potential - theoretic considerations which also give a form of the second main theorem for holomorphic curves @xcite .
a survey of other results obtained with this method is @xcite . to prove our main result we combine the method of @xcite with two new ingredients .
the first one is a quantitative form of the uniqueness theorem for harmonic functions in the form of a generalization of the three circles theorem of hadamard by nadirashvili @xcite .
the second ingredient is theorem [ derivthm ] , which gives an estimate for @xmath43 in terms of the fubini
study area of the image of @xmath3 , in the case that @xmath3 omits @xmath9 hyperplanes in general position .
this estimate seems to be new even for @xmath44 , @xcite .
the plan of this paper is as follows . in section [ linalgsec ]
we recall some formulas for the fubini - study derivative and some of its basic properties . in section [ rieszjensensec ] , we recall the riesz theorem and jensen s formula . in section [ areatoderivativesec ] , we show using the poisson formula and harnack s inequality that if a holomorphic map @xmath3 from a disc to @xmath1 omits @xmath9 hyperplanes in general position and covers finite area @xmath45 measured in the fubini - study metric , then @xmath43 can be effectively bounded in terms of @xmath46 the dimension @xmath11 and the omitted hyperplanes . in section [ cartansec ] , we recall cartan s lemma in a form we will need . in section [ nadirashvilisec ] ,
we state nadirashvili s generalization of the three circles theorem . in section [ coveringsec ]
, we recall a covering lemma of rickman that was also used in @xcite .
finally , in section [ mainsec ] , we put the ingredients together to give our landau and schottky theorems .
_ acknowledgment .
_ work on this article began when both authors visited the christian albrechts universitt zu kiel with financial support from the humboldt foundation and continued while the first author was a member at msri .
the authors would like to thank these institutions for their financial support and hospitality .
if @xmath47 are @xmath9 linear forms defining @xmath9 hyperplanes in general position in @xmath1 normalized as in the introduction so that @xmath48 then for any point @xmath49 $ ] in @xmath1 , we have @xmath50 where @xmath51 and @xmath52 are defined as in equation @xmath53 .
let @xmath54 $ ] be a holomorphic map from a domain in @xmath55 to @xmath1 given by homogeneous coordinate functions @xmath56 which are holomorphic without common zeros .
let @xmath57.$ ] we recall the following formulas for the fubini - study derivative @xmath5 of @xmath58 @xmath59}{\ds\| f\|^4 } \;=\;\frac{\ds\sum_{j=0}^n\sum_{k=0}^n|f_j|^2|f_k|^2 \frac{\bar f_j'}{\bar f_j } \left[\frac{f_j'}{f_j}-\frac{f_k'}{f_k } \right]}{\ds\| f\|^4}\\ \label{fsiii } & = \ds \frac{\ds\sum_{j=0}^{n-1}\sum_{k = j+1}^n |f_jf_k'-f_kf_j'|^2}{\ds\| f\|^4}\\ \noalign{\vskip 6pt } \label{fsiv } & = \ds \frac{\ds||f\wedge f'||^2}{\ds||f||^4}.\end{aligned}\ ] ] [ changecoordprop ] let @xmath60 be @xmath9 hyperplanes in general position in @xmath1 given by homogeneous forms normalized so that the coefficient vectors have norm @xmath61 let @xmath54 $ ] be a holomorphic curve in @xmath1 with homogeneous coordinate functions @xmath62 let @xmath63 $ ] be the holomorphic curve in @xmath1 given by @xmath64 then , @xmath65 where @xmath66 and @xmath67 are defined as in equation ( [ eigenvaleqn ] ) .
only the left - hand inequality involving @xmath66 will be used in the sequel . with notation as defined preceding equation ( [ eigenvaleqn ] )
, by linear algebra , @xmath68 and hence the proposition follows from ( [ fsiv ] ) .
in this section we recall the riesz theorem and the jensen formula , and we derive a corollary of the jensen formula that says a subharmonic function with large riesz mass can not stay close to a harmonic function .
let @xmath69 be a @xmath70 function on a domain in @xmath71 we recall the differential operator @xmath72 which can be defined by @xmath73 where @xmath74 is the lebesgue area form on @xmath71 one of the advantages of the @xmath72 notation is that the factors @xmath75 and @xmath76 do not appear in fundamental formulas .
as usual the operator @xmath72 is extended to the space @xmath77 of schwartz distributions .
when @xmath69 is subharmonic , @xmath78 defines a ( locally finite , positive ) borel measure known as the riesz measure of @xmath79 [ rieszthm ] let @xmath69 be subharmonic on the unit disc @xmath80 with finite riesz mass .
then , @xmath81 is harmonic on @xmath82 [ greenjensen ] let @xmath83 be a difference of two subharmonic functions on @xmath84 and let @xmath85 then , @xmath86.\ ] ] [ jensencor ] let @xmath69 be a subharmonic function on @xmath84 and let @xmath87 be harmonic on @xmath88 let @xmath89 and assume @xmath90 then @xmath91 on the one hand by ( a ) and the jensen formula , @xmath92 = \int_r^r \frac{dt}{t}\int\limits_{|z|\le t}dd^c v\ge \log \frac{r}{r}.\ ] ] on the other hand by ( b ) since @xmath87 is harmonic , @xmath93 \frac{d\theta}{2\pi } \le\frac{1}{2}\int_0^{2\pi}[u(re^{i\theta } ) -u(re^{i\theta})+2\varepsilon]\frac{d\theta}{2\pi } \le \varepsilon . \eqno{\qedhere}\ ] ]
for a holomorphic curve omitting @xmath0 hyperplanes in general position , our eventual goal is to bound @xmath43 explicitly in terms of the @xmath0 given omitted hyperplanes .
what we will first bound is the integral of @xmath94 on a disc centered at the origin . in this section ,
we see how to obtain a derivative bound from such an area bound .
this section is based on @xcite and appears to be new even in dimension one .
[ logderivprop ] let @xmath3 be a holomorphic function without zeros on @xmath35 such that @xmath95 then , @xmath96 let @xmath97 since @xmath98 differentiating the poisson formula for @xmath99 gives us @xmath100 because the statement follows letting @xmath101 [ derivthm ] let @xmath54 $ ] be a holomorphic map from the unit disc @xmath35 to @xmath1 which omits the @xmath9 hyperplanes in general position .
if @xmath102 then @xmath103.\ ] ] before giving a proof , we make some comments .
the above theorem is really only of interest when @xmath104 when @xmath105 and @xmath38 a theorem of dufresnoy @xcite ( or see also ( * ? ? ?
6.1 ) or ( * ? ?
2.8.3 ) ) says @xmath106 ^ 2 \le \frac{\sigma}{1-\sigma},\ ] ] even without the assumption that @xmath3 is zero free .
the linear functions @xmath107 show the dufresnoy result is sharp in that setting .
the same example shows that one can not remove the assumption on the omitted hyperplanes from theorem [ derivthm ] .
note also that our estimate does not even tend to @xmath108 as @xmath109 we briefly comment on the sharpness of the coefficient in front of @xmath110 consider the case @xmath44 and a zero free holomorphic function .
let @xmath111 then , @xmath112 and because @xmath113 wraps the disc around the sphere so as to cover @xmath114 hemispheres , @xmath115 hence , @xmath116 no matter how large @xmath114 is , and thus the @xmath117 we obtain in front of @xmath45 is at worst @xmath118 times too big . by first working on a disc of radius @xmath119 and then taking a limit
as we may assume without loss of generality that @xmath3 is analytic on by abuse of notation , let @xmath18 also denote the linear forms defining the hyperplanes , and normalize them so that @xmath120 let and without loss of generality , assume the least harmonic majorant of the @xmath121 on the unit disc @xmath122 is @xmath108 and that let and let then @xmath87 and @xmath69 are subharmonic and satisfy @xmath123 let @xmath124 and @xmath125 because @xmath126 we have @xmath127 by proposition [ changecoordprop ] , formula ( [ fsii ] ) , proposition [ logderivprop ] and our assumption that @xmath128 thus , it suffices to bound @xmath129 let @xmath130 from the jensen formula and the assumed bound on the integral of @xmath131 we have @xmath132 = \int_r^1 \frac{dt}{t}\int_{|z|\le t } ( f^\#)^2\frac{da}{\pi } \le \sigma \log \frac{1}{r}.\ ] ] thus , there is some point @xmath133 with @xmath134 so that @xmath135 where the outside inequalities follow from ( [ derivthmstar ] ) , ( [ derivthmstarstar ] ) , and the fact that @xmath136 when @xmath137 let @xmath138 be the index such that @xmath139 and then apply harnack s inequality to conclude @xmath140.\end{aligned}\ ] ] taking @xmath141 we get @xmath142.\qedhere\ ] ]
in this section we recall cartan s lemma , a well - known fundamental estimate whose significance in the study of function theory was recognized by bloch and made rigorous by cartan .
[ blashthm ] let @xmath143 be a finite borel measure on @xmath144 and let @xmath145 be the blaschke potential @xmath146 let @xmath147 let @xmath148 and let @xmath149 then , there is a countable family of exceptional discs @xmath150 the sum of whose radii does not exceed @xmath151 such that for those @xmath152 with @xmath153 and not in any of the @xmath150 we have @xmath154 where @xmath155 theorem [ blashthm ] is stated and proved for finite sums in cartan @xcite , which is also reproduced in ( * ? ?
viii.3.3 ) . the argument for general @xmath143 with finite
mass is the same , so we omit the proof here . [ cartancor ]
let @xmath69 be a negative subharmonic function on @xmath35 such that for all @xmath156 @xmath157 let @xmath158 @xmath159 @xmath160 and @xmath161 be as in the theorem .
then , for all @xmath153 outside of a countable set of discs the sum of whose radii is at most @xmath151 we have @xmath162 note that the radius of each exceptional disc is at most @xmath158 and hence the collective area of the exceptional discs is at most @xmath163 in ( * ? ? ?
v ) , cartan states this result in the case @xmath164 for an analytic function @xmath165 with @xmath166 but with a slightly worse expression for @xmath167 cartan s proof only makes use of theorem [ blashthm ] in the case of a discrete measure , rather than for a general measure as we shall do . if @xmath168 the estimate is trivial , so we assume @xmath169 is finite .
let @xmath170 be a large constant so that @xmath171 and let @xmath172 note that @xmath173 is subharmonic , that @xmath174 for @xmath175 by the assumption that @xmath176 and that @xmath177 by the riesz theorem , @xmath178 is harmonic on @xmath82 however , both @xmath173 and the integral potential vanish identically for @xmath179 and so @xmath180 is actually equal to the integral .
we then apply the theorem to conclude the existence of exceptional discs as in the statement of this corollary such that for those @xmath152 outside those discs and @xmath181 @xmath182 we thus conclude that for these same @xmath183 we have @xmath184
in this section we state a result of nadirashvili @xcite specialized to dimension 2 .
we slightly reformulate nadirashvili s statement to get explicit constants convenient for our application .
[ threecircthm ] let @xmath87 be a harmonic function in the unit disc with @xmath185 let and assume for @xmath186 and some @xmath187 then , for @xmath188 as in ( * ? ? ?
2.1 ) , the classical hadamard three circles theorem implies , @xmath189 ^ 2\frac{d\theta}{2\pi } \le \rho^{2\tau } \qquad r < \rho < 1.\ ] ] thus , @xmath190 ^ 2 \frac{d\theta}{2\pi}d\rho \le \frac{(2r)^{2\tau+2}-r^{2\tau+2}}{\tau+1 } \le ( 2r)^{2\tau+2}-r^{2\tau+2}.\ ] ] trivially , @xmath191 ^ 2 \frac{d\theta}{2\pi}d\rho \le r^{2\tau+2}.\ ] ] let @xmath160 be a point with @xmath192 such that for then , by jensen s inequality , @xmath193 ^ 2 \frac{d\theta}{2\pi}d\rho\\ & \ge \int_0^r 2 t \int_0^{2\pi}[u(z_0+te^{i\theta})]^2 \frac{d\theta}{2\pi}dt\\ & \ge \int_0^r2t\left[\int_0^{2\pi}u(z_0+te^{i\theta } ) \frac{d\theta}{2\pi}\right]^2dt = [ u(z_0)]^2r^2 . \qedhere\end{aligned}\ ] ] [ remez ] let @xmath194 be a disc of radius @xmath195 in @xmath55 and let @xmath196 be a subset of @xmath194 with area @xmath197 let @xmath198 be a real polynomial of degree @xmath199 on @xmath200 and assume for @xmath152 in @xmath201 then , for all @xmath152 in @xmath202 @xmath203 see ( * ? ? ?
* lem . 2 ) .
if @xmath204 is a complex number and @xmath205 let @xmath206 denote the disc of radius @xmath195 centered at @xmath207 [ nadthm ] let @xmath87 be a harmonic function in the unit disc such that @xmath185 let let @xmath196 be a subset of @xmath208 with area @xmath209 let @xmath210 and assume @xmath211 if @xmath212 for all @xmath152 in @xmath213 then there is a point @xmath160 in @xmath208 such that @xmath214 moreover , @xmath215 } + \log 2 \qquad\textit{for~all~}|z|<\rho.\ ] ] let @xmath160 be a point in @xmath208 such that @xmath216 such a point exists because @xmath208 is covered by at most @xmath217 discs of radius @xmath195 with centers on a square lattice .
let @xmath218 and let @xmath219 be the @xmath199-th taylor polynomial of @xmath220 because by estimating the error in taylor s theorem using @xmath221 we know that for @xmath152 in @xmath222 we have @xmath223 thus , @xmath224 for @xmath152 in @xmath225 by proposition [ remez ] , if @xmath152 is in @xmath222 then @xmath226 where the last equality follows from our choice of @xmath227 this together with ( [ tayloreqn ] ) gives ( [ nadone ] ) .
let @xmath228 then , @xmath69 is harmonic on the unit disc , @xmath229 and @xmath230 let @xmath231 be such that @xmath232 then , we conclude by theorem [ threecircthm ] that @xmath233 because @xmath234 we conclude that @xmath235 which gives ( [ nadtwo ] ) .
this section describes a covering lemma of rickman @xcite as corrected in @xcite .
we provide a proof here to get a slightly better constant .
if @xmath204 is a complex number and @xmath205 let @xmath206 denote the disc of radius @xmath195 centered at @xmath207 for a complex number @xmath236 with @xmath237 and an integer @xmath238 denote by @xmath239 [ covprop ] let @xmath204 be a complex number with @xmath240 and for @xmath241 let @xmath242 where @xmath243 is understood to be @xmath244 if @xmath245 then @xmath246 without loss of generality , assume @xmath247 the proof is illustrated by figure [ circfig ] .
the dotted circle in the left - hand picture represents the boundary of @xmath248 and the bold circle is the boundary of @xmath249 the picture on the right is a magnification of the bold circle on the left with only three of the covering discs of the form @xmath250 drawn in .
the point is that all the covering discs have radius at least as big as @xmath251 thus , it suffices to show that the sector of the ring ( shaded in the figure on the right ) @xmath252 is contained in the disc @xmath253 this is true because the four corner points of the annular sector are easily seen to be contained in @xmath254 using the law of cosines on the triangles that have as vertices @xmath255 @xmath256 and one of the corner points of the sector , noting that the side lengths of the triangle and @xmath257 vary in direct proportion with @xmath258 as @xmath204 varies .
[ rickmanlemma ] let @xmath143 be a borel measure on @xmath35 such that @xmath35 has finite @xmath143-measure .
let @xmath259 be an integer and let @xmath260 then , there is a complex number @xmath204 with @xmath240 such that @xmath261 and @xmath262 choose @xmath204 with @xmath240 such that @xmath263 which immediately implies the second inequality of the lemma by considering @xmath264 the first inequality follows by iterating proposition [ covprop ] .
we now prove our upper bound on the fubini - study area covered by a holomorphic curve omitting @xmath0 hyperplanes in general position in @xmath17 [ mainthm ] let @xmath3 be a holomorphic map from the unit disc @xmath35 to @xmath1 omitting @xmath0 hyperplanes @xmath29 in general position .
then , @xmath265 where @xmath32 is the binomial coefficient @xmath266 and @xmath30 where @xmath51 and @xmath52 are defined in equation @xmath53 . by a standard limiting argument we may assume @xmath3 is holomorphic in a neighborhood of @xmath267 let @xmath268 $ ] be projective coordinate functions for @xmath269 let @xmath270 and for @xmath271 let @xmath272 where as usual we also use @xmath18 to denote the linear forms defining the hyperplanes normalized so @xmath120 for @xmath273 by lemma [ rickmanlemma ] , there exists @xmath204 and @xmath195 such that @xmath274 and such that @xmath275 let @xmath276 be the least harmonic majorant of @xmath277 on @xmath278 define @xmath279 \qquad\textnormal{and}\qquad v_j(z)=\frac{1}{\sigma_{a , r}}[u_j(a+rz)-\tilde u(a+rz)].\ ] ] then @xmath280 are harmonic and negative on @xmath281 with @xmath282 when @xmath283 the function @xmath69 is subharmonic on @xmath284 with @xmath285 let @xmath291 be an index set of cardinality @xmath292 and let @xmath293 by @xmath294 because we have @xmath0 hyperplanes total , at every point @xmath160 with @xmath295 there are ( at least ) @xmath9 distinct indices @xmath296 such that @xmath297 and hence there is at least one such index set @xmath298 such that @xmath299 by re - ordering the indices , without loss of generality , we will assume @xmath298 contains @xmath300 from now on we will consider only these indices , and we will denote @xmath301 simply by @xmath201 by shrinking @xmath196 if necessary , we may assume @xmath302 by the jensen formula and @xmath303 @xmath304 = \int_2^{16 } \frac{dt}{t}\int\limits_{|z|\le t } dd^c v \le 16 ^ 4\covconst\log 8.\ ] ] because @xmath305 for @xmath306 we conclude that there is a point @xmath307 with @xmath308 such that @xmath309 we now apply corollary [ cartancor ] to @xmath310 with to conclude that @xmath311\ ] ] for @xmath312 and outside a collection of discs whose radii @xmath313 are such that @xmath314 as @xmath315 this implies the exceptional discs have collective area less than the area of @xmath213 and so there is a point @xmath160 in @xmath196 with @xmath316 \ge \cartanconst(1/8,\eta)[-2\cdot16 ^ 4\covconst\log8 - 2\varepsilon].\ ] ] for @xmath317 because @xmath280 is within @xmath318 of @xmath69 at @xmath319 we conclude @xmath320.\ ] ] applying the harnack inequality twice implies @xmath321 letting @xmath322 we get for @xmath323 and @xmath324 now suppose there is a constant @xmath325 such that for any two indices @xmath138 and @xmath326 with @xmath327 we have @xmath328 because at every point @xmath183 one of these @xmath9 functions @xmath280 comes within @xmath318 of @xmath83 and they are all within @xmath329 of each other , we have that @xmath330 we can now apply corollary [ jensencor ] with @xmath331 and @xmath332 to conclude @xmath333 we then consider the case that and so from the above , @xmath334 in this case , there are indices @xmath138 and @xmath326 and a point @xmath335 with @xmath336 such that @xmath337 now , @xmath338 is harmonic on @xmath339 @xmath340 on @xmath213 and @xmath341 apply theorem [ nadthm ] to @xmath342 which is bounded by @xmath343 on @xmath213 to conclude from ( [ nadtwo ] ) that @xmath344 letting @xmath345 and simplifying we get , @xmath346 because this is much greater than @xmath347 there was no harm in our considering only @xmath348 combining the above with ( [ onequarter ] ) and recalling that we conclude @xmath349 and hence the theorem follows .
[ landauthm ] let @xmath29 be @xmath0 hyperplanes in general position in @xmath17 let @xmath30 and @xmath31 let @xmath32 be the binomial coefficient @xmath33 and let @xmath34 let @xmath3 be a holomorphic map from the unit disc @xmath35 to @xmath1 omitting @xmath36 then for @xmath144 @xmath350 when @xmath38 landau s original theorem gave a bound on @xmath351 in terms of @xmath352 for analytic functions on the unit disc omitting @xmath108 and @xmath61 in this context , @xmath353 and so theorem [ landauthm ] also gives a bound on @xmath351 in terms of @xmath354 of course , the bound is non - optimal in this special case .
as we discussed in the introduction , the kobayashi metric on the complement of the hyperplanes @xmath18 is a metric , so theorem [ landauthm ] can be interpreted as a global lower - bound over this hyperplane complement on the ratio of the infinitesimal kobayashi metric to the infinitesimal fubini - study metric .
unfortunately , our method does not give any information in terms of the geometry of the omitted hyperplanes on where the minimum ratio occurs .
for geometrically symmetric hyperplane configurations , one could hope to determine these minimum points .
this was done in the one - dimensional case in @xcite and @xcite .
note that the dependence on @xmath355 in theorem [ landauthm ] is essentially best possible .
indeed , consider @xmath44 and consider the points @xmath356 @xmath114 and @xmath357 for @xmath114 large .
then the map @xmath358 omits the three points provided @xmath359 then , @xmath360 approaches @xmath361 as @xmath195 approaches @xmath362 because we have included @xmath108 and @xmath357 among our three points , @xmath363 in this case .
a straightforward computation shows that @xmath364 an easy application of lhpital s rule also shows that @xmath365 and hence the linear appearance of @xmath355 in the inequality in theorem [ landauthm ] is correct .
of course the enormous constant in front is not optimal .
now consider the case where the three points are @xmath366 @xmath367 and @xmath357 for @xmath114 large . in this case
bonk and cherry @xcite gave a sharp upper bound on @xmath368 if @xmath369 denotes the universal covering map of which sends @xmath108 to @xmath356 then the map @xmath370 has the largest spherical derivative at the origin in comparison with all maps omitting @xmath366 @xmath371 and @xmath372 one can even compute using schwarz triangle functions that @xmath373 one sees as above that here @xmath355 is asymptotic to @xmath374 as @xmath375 and that @xmath376 is asymptotic to @xmath366 and thus , the hyperplane dependence through the term @xmath376 is also not too bad .
the appearance of the combinatorial coefficient @xmath32 in the exponent of our estimate causes the estimate to deteriorate rather severely as the dimension @xmath199 increases .
it would be interesting to find out if there must be some dependence on @xmath199 in the estimate beyond the fact that @xmath355 depends implicitly on @xmath377 the appearance of this combinatorial constant also prevents our method from giving a better bound when @xmath3 omits more than @xmath0 hyperplanes . if @xmath3 omits more than @xmath0 hyperplanes , the best estimate our method gives is to choose the @xmath0 hyperplanes that give the best @xmath355 from among the omitted hyperplanes .
this is clearly a deficiency in our approach . for maps to @xmath1 omitting @xmath378 hyperplanes
, it would be nice to prove an estimate that improves as @xmath378 increases , as in ( * ? ? ?
5.7.4 ) when @xmath379 let @xmath45 be the upper bound on @xmath380 obtained from theorem [ mainthm ] .
we apply theorem [ derivthm ] taking the best choice of @xmath9 hyperplanes among our given @xmath0 omitted hyperplanes to conclude that @xmath381 \le 11\cdot32\sigma,\ ] ] since @xmath382 is clearly less than @xmath46 and the factor @xmath383 in front of @xmath45 comes from rescaling @xmath384 to the unit disc before applying theorem [ derivthm ] .
[ schottkythm ] with the hypotheses and notation of theorem [ landauthm ] , let @xmath387 recalling that we have normalized the defining forms of our hyperplanes so that @xmath120 then for @xmath323 and @xmath144 @xmath388.\ ] ] mmmmm v. babets , pseudoforms with negative curvature on @xmath389 ( russian ) , izv .
nauk sssr ser.mat .
* 50 * ( 1986 ) , 13261337 , 1344 ; translation in math .
ussr - izv .
* 29 * ( 1987 ) , 677688 .
v. babets , pseudoforms with negative curvature on @xmath389 ii ( russian ) , izv .
nauk ser . mat .
* 57 * ( 1993 ) , 132138 ; translation in russian acad .
* 43 * ( 1994 ) , 119125 .
m. bonk and w. cherry , bounds on spherical derivatives for maps into regions with symmetries , j. anal .
* 69 * ( 1996 ) , 249274 .
m. bonk and w. cherry , metric distortion and triangle maps , ann .
* 24 * ( 1999 ) , 489510 .
h. cartan , sur les systmes de fonctions holomorphes varits linaires lacunaires , ann .
sci . ecole norm .
* 45 * ( 1928 ) , 255346 .
w. cherry and z. ye , _
nevanlinna s theory of value distribution . the second main theorem and its error terms , _ springer - verlag , 2001 . m. cowen , `` the kobayashi metric on @xmath390 hyperplanes '' in _ value distribution theory , part a , _ dekker , 1974 , pp .
m. cowen , the method of negative curvature : the kobayashi metric on @xmath391 minus @xmath392 lines , trans .
* 319 * ( 1990 ) , 729745 .
j. dufresnoy , sur les domaines couvertes par les valeurs dune fonction mromorphe algebrode , ann .
sci . ecole norm .
( 3 ) * 58 * ( 1941 ) , 179259 . j. dufresnoy , thorie nouvelle des familles complexes normales
. applications ltude des fonctions algbrodes , ann .
sci . ecole norm .
( 3 ) * 61 * ( 1944 ) , 144 .
a. eremenko , a picard type theorem for holomorphic curves , period .
* 38 * ( 1999 ) , 3942 .
a. eremenko , an estimate for spherically @xmath393-valent functions , preprint 1998 , www.math.purdue.edu/~eremenko/newprep.html a. eremenko , value distribution and potential theory , proceedings of the international congress of mathematicians , vol .
2 , p. 681 - 690 , higher education press , beijing , 2002 .
a. eremenko and m. sodin , distribution of values of meromorphic functions and meromorphic curves from the standpoint of potential theory ( russian ) , algebra i analiz * 3 * ( 1991 ) , 131164 ; translation in st .
petersburg math .
j. * 3 * ( 1992 ) , 109136 .
p. hall , landau and schottky theorems for holomorphic curves , michigan math .
j. * 38 * ( 1991 ) , 207223 . w. hayman , _ meromorphic functions , _ clarendon press , 1964
. w. hayman , _ subharmonic functions , _ volume ii , london mathematical society monographs * 20 * , academic press , 1989 .
j. a. hempel , the poincar metric on the twice punctured plane and the theorems of landau and schottky , j. london math
. soc .
( 2 ) * 20 * ( 1979 ) , 435445 .
j. a. jenkins , on explicit bounds in landau s theorem ii , canad .
j. math .
* 33 * ( 1981 ) , 559562 .
s. kobayashi , _ hyperbolic complex spaces , _ grundlehren der mathematischen wissenschaften * 318 * , springer - verlag , 1998
. j. korevaar and j. meyers , logarithmic convexity for supremum norms of harmonic functions , bull .
london math .
* 26 * ( 1994 ) , 353362 .
e. landau , ber eine veralgemeinerung des picardschen satzes , s .- b .
preu akad .
wiss . , phys .- math .
( 1904 ) , 1118 - 1133 .
s. lang , _ introduction to complex hyperbolic spaces , _ springer - verlag , 1987 .
n. nadirashvili , a generalization of hadamard s three circles theorem ( russian ) , vestnik moskov . univ .
* 31 * ( 1976 ) , 3942 ; translation in moscow univ .
math . bull .
* 31 * ( 1976 ) , 3032 .
s. rickman , on the number of omitted values of entire quasiregular mappings , j. danalyse math .
* 37 * ( 1980 ) , 100 - 117 . | we prove an effective version of a theorem of dufresnoy : for any set of @xmath0 hyperplanes in general position in @xmath1 , we find an explicit constant @xmath2 such that for every holomorphic map @xmath3 from the unit disc to the complement of these hyperplanes , we have @xmath4 , where @xmath5 denotes the norm of the derivative measured with respect to the fubini - study metric .
this result gives an explicit lower bound on the royden function , _
i.e. , _ the ratio of the kobayashi metric on the hyperplane complement to the fubini - study metric . our estimate is based on the potential - theoretic method of eremenko and sodin . |
to completely describe the lattice of subalgebras @xmath2 of a finite algebra @xmath0 is a difficult task .
moreover , it is not obvious how to store the information about @xmath2 efficiently , as the cardinality and complexity of @xmath2 is typically much larger than that of @xmath0 .
fortunately , sometimes there is a procedure that allows us to calculate the join @xmath3 and meet @xmath4 for every @xmath5 , @xmath6 .
for instance , it is easy to find the join and meet in any boolean algebra @xmath0 , although @xmath7 grows exponentially in @xmath8 .
it is this ability to calculate @xmath9 and @xmath10 that is often understood as a complete description of @xmath2 .
most of the time we are not so lucky , though , and there is no apparent way to find joins and meets .
the main reason is that @xmath3 and @xmath4 can be far from both @xmath5 and @xmath11 in the lattice @xmath2 .
it is therefore more convenient to have access to a procedure that gives a complete _ local _ description of @xmath2 . assuming that it is possible to find all maximal subalgebras of @xmath12 and all subalgebras @xmath13 in which @xmath5 is maximal , the lattice @xmath2 can be built up inductively .
we will refer to all subalgebras immediately above and immediately below @xmath5 in @xmath2 as _ neighbors _ of @xmath5 , and we denote the set they form by @xmath14 . in this context , it is worth paying attention to the automorphism group @xmath1 and its natural action on @xmath0 , since the neighborhoods of @xmath5 and @xmath11 will be `` the same '' for @xmath5 , @xmath6 belonging to the same orbit of transitivity of @xmath1 .
thus , the lattice @xmath2 can be fully described as long as we find 1 .
one representative @xmath5 from each orbit of @xmath1 , 2 .
the neighborhood @xmath14 for every representative @xmath5 , and 3 .
an automorphism of @xmath0 mapping @xmath11 onto @xmath5 , for every representative @xmath5 and every @xmath11 from the orbit of @xmath5 . to save space , we can store subalgebras by their generating sets , and substitute @xmath14 and the automorphisms required by @xmath15 with an efficient algorithm producing those .
the purpose of this paper is twofold .
first , to introduce a general tool
hasse constants that is of some help in all three tasks @xmath16 , @xmath17 , @xmath15 . secondly , to use hasse constants to describe the subloop lattice of the smallest nonassociative simple moufang loop @xmath18 .
the investigation of @xmath19 occupies most of this paper , and is inevitably of rather detailed nature .
we maintain that the power of hasse constants is sufficiently demonstrated by the fact that @xmath19 was not known before ( see acknowledgement ) , especially given the importance of @xmath18 for the real octonions .
although considerable invention will be required in each particular case , we believe that hasse constants will help to keep track in investigation of any subalgebra lattice . a word about the notation : we write @xmath20 for the cyclic group of order @xmath21 , @xmath22 for the dihedral group of order @xmath23 , and @xmath24 for the elementary abelian @xmath25-group of order @xmath26 . a subalgebra generated by the set @xmath27
well be denoted by @xmath28 .
let @xmath5 , @xmath11 , @xmath0 be finite ( universal ) algebras , @xmath12 . for @xmath29 , let @xmath30 denote the orbit of @xmath31 under the natural action of @xmath1 on the set of subalgebras of @xmath0 isomorphic to @xmath31 .
we will speak of the subalgebras of @xmath0 isomorphic to @xmath31 as _ copies _ of @xmath31 in @xmath0 . define @xmath32 furthermore , when @xmath13 , let @xmath33 in words , @xmath34 counts the number of copies of @xmath11 in @xmath0 , @xmath35 counts the number of copies of @xmath11 in @xmath0 containing @xmath5 , and @xmath36 counts the number of copies of @xmath11 in @xmath0 containing @xmath5 and in the same orbit as @xmath11 . yet
another description of these constants is perhaps the most appealing . for @xmath13 , the constant @xmath34 counts the number of edges connecting @xmath0 to a copy of @xmath11 in the complete hasse diagram of @xmath2 .
the remaining two constants can be interpreted in a similar way
. we will therefore refer to them jointly as _
hasse constants_. note that @xmath37 if @xmath1 acts transitively on the copies of @xmath11 in @xmath0 .
[ lm : counting ] let @xmath5 , @xmath11 , @xmath0 be algebras , @xmath12 . 1 .
if @xmath38 , @xmath39 , then @xmath40 .
2 . if @xmath41 , @xmath38 , then @xmath42 .
3 . if @xmath41 , @xmath13 , @xmath43 , then @xmath44 . part ( i ) is obvious from the definition of @xmath34 .
the equality @xmath45 holds if @xmath46 .
let @xmath41 , and let @xmath47 be an automorphism mapping @xmath5 to @xmath48 . then @xmath49 .
this proves ( ii )
. part ( iii ) is similar ( use @xmath50 ) .
this example shows that the constants @xmath35 , @xmath51 may differ even though @xmath52 .
let @xmath0 be the group @xmath53 , @xmath54 , @xmath55 , @xmath56 , @xmath25 , @xmath57 , and denote by @xmath58 the two - element subgroup of @xmath59 .
the lattice of subgroups of @xmath0 is depicted in figure @xmath60 . with @xmath61
, we have @xmath62 .
[ pr : counting ] let @xmath0 be an algebra , @xmath5 , @xmath13 .
let @xmath63 , @xmath64 , @xmath65 be representatives from all orbits @xmath66 , @xmath64 , @xmath67 of the action of @xmath1 on the copies of @xmath5 in @xmath0 .
similarly , let @xmath68 , @xmath64 , @xmath69 be representatives for @xmath11 .
then @xmath70 when @xmath1 acts transitively on the copies of @xmath11 @xmath71i.e .
, when @xmath72 , then @xmath73 coincides with @xmath74 .
when @xmath1 acts transitively on the copies of @xmath5 @xmath71i.e .
, when @xmath75 , then @xmath76 since every copy of @xmath11 in @xmath0 belongs to exactly one orbit @xmath77 , @xmath78 follows . to establish @xmath73 ,
count twice the cardinality @xmath79 of @xmath80 on the one hand , @xmath81 on the other hand , @xmath82 the proof of @xmath74 is similar to @xmath73 . just count twice the cardinality of the set @xmath83 when @xmath84 , @xmath85 and @xmath86 follow immediately from @xmath73 and @xmath74 , respectively .
loops satisfying one of the equivalent _ moufang identities _ , for instance the identity @xmath88 are called _ moufang loops _ @xcite . by a result of kunen @xcite ,
every quasigroup satisfying @xmath89 is a moufang loop . obviously , every group is a moufang loop .
moufang loops are _ power associative _
( i.e. , every @xmath56-generated subloop is a group ) , in fact _ diassociative _ ( i.e. , every @xmath25-generated subloop is a group ) .
every element @xmath90 of a moufang loop has a ( unique ) two sided inverse @xmath91 .
paige @xcite , doro @xcite and liebeck @xcite showed that there is only one class of nonassociative finite simple moufang loops , consisting of loops @xmath92 , one for each finite field @xmath93 .
these loops are best studied via composition algebras . following @xcite ,
let @xmath94 be the unique split octonion algebra over @xmath93 .
then @xmath92 is isomorphic to the multiplicative loop of elements of norm @xmath56 in @xmath94 modulo the center .
the algebra @xmath94 was first constructed by zorn as follows .
given a prime power @xmath95 , let @xmath96 be the standard dot product and @xmath97 the standard vector product on @xmath98 .
then the algebra of _ vector matrices _ @xmath99 with addition defined entry - wise and multiplication governed by @xmath100 is isomorphic to @xmath94 .
the norm on @xmath94 coincides with the determinant @xmath101 .
the neutral element is @xmath102 and every element @xmath90 with nonzero norm has inverse @xmath103 the order of @xmath92 is @xmath104 when @xmath95 is even , and @xmath105 when @xmath95 is odd @xcite .
notably , the loop @xmath18 has also connections to the standard real ( division ) octonion algebra .
namely , it is isomorphic to the integral real octonions of norm @xmath56 modulo the center ( cf .
@xcite , @xcite ) .
let us recall loops of type @xmath106 , first constructed in @xcite .
for a group @xmath107 of order @xmath21 , define new multiplication @xmath96 on @xmath108 by @xmath109 then @xmath110 is a moufang loop , and we denote it by @xmath111 .
it is nonassociative if and only if @xmath107 is nonabelian ( cf.@xcite ) .
write @xmath112 for some element @xmath113 .
lemma [ lm : structuralm2n ] ( cf .
* prop 4.12 ) ) is easy to prove once you realize that 1 .
every element of @xmath114 is of order @xmath25 , 2 .
@xmath115 , @xmath116 , 3 .
every subloop @xmath117 of @xmath111 satisfies @xmath118 .
[ lm : structuralm2n ] let @xmath107 be a group of order @xmath21 , and let @xmath112 be constructed as above . 1
. we have @xmath119 2 .
@xmath120 for every @xmath121 , @xmath122 , @xmath123 , @xmath124 .
3 . for @xmath125 ,
@xmath126 4 .
@xmath127 for every @xmath128 , @xmath129 with @xmath130 .
5 . when @xmath131 and @xmath132 , then @xmath133 .
it was proved in @xcite that @xmath111 is presented ( in the variety of moufang loops ) by @xmath134 whenever @xmath107 is a @xmath25-generated group with presentation @xmath135 in particular , presentations for the loops @xmath136 and @xmath137 can be obtained from this result .
we proceed to describe the subloop lattice of @xmath18 , guided by steps @xmath16 , @xmath17 , @xmath15 of section [ sc : introduction ] .
we fulfil @xmath16 and give reasonable amount of details with respect to @xmath17 and @xmath15 . in particular , we calculate all hasse constants @xmath34 , @xmath35 , @xmath36 , for @xmath138 . at several places , the reader will be kindly asked to verify a few details by straightforward , easy calculations .
most of these calculations are reduced to a quick glance into table [ tb : main ] .
the table itself can be checked for accuracy within minutes , using lemma [ lm : involutions ] .
no machine computation is needed .
fix @xmath139 and @xmath140 .
it is easy to see that @xmath0 consists of @xmath141 elements of order @xmath56 , @xmath25 , @xmath142 .
more precisely , @xmath143 satisfies @xmath144 if and only if @xmath145 and @xmath146 ; and @xmath147 if and only if @xmath148 .
to linearize our notation , we write @xmath149}_{a}}$ ] when @xmath144 , and @xmath150 when @xmath147 . since @xmath151 can be calculated from @xmath152 , @xmath153 when @xmath144 , we further simplify involutions to @xmath154}$ ] . note that @xmath155 , where the addition is modulo @xmath25 .
elementary counting reveals that there are @xmath156 involutions and @xmath157 elements of order @xmath142 in @xmath0 . using the language of hasse constants , @xmath158 , @xmath159 .
chein classified all nonassociative moufang loops of order at most @xmath156 @xcite , and we will call such moufang loops _ small_. since @xmath0 has @xmath141 elements , every proper subloop of @xmath0 is small and can be found in chein s list . as in @xcite
, we say that a finite loop @xmath160 has the _ weak cauchy property _ when it contains a subloop of order @xmath161 for every prime @xmath161 dividing @xmath162 .
it has the _ weak lagrange property _ if @xmath163 divides @xmath164 for every @xmath165 .
finally , @xmath160 has the _ strong cauchy _ ( _ lagrange _ ) _ property _ if every subloop of @xmath160 has the weak cauchy ( lagrange ) property .
since @xmath166 divides @xmath8 , @xmath0 does not have the weak cauchy property .
however , it follows from ( * ? ? ?
xiv ) that all small moufang loops have it .
( they also have the strong lagrange property .
if one proves that every @xmath92 has the strong lagrange property , it will follow that all moufang loops have it ( cf . @xcite ) .
as of now , this is an open question .
glauberman proved @xcite that all moufang loops of odd order have the strong lagrange property . )
[ cr : restrictedorders ] the order of every proper subloop of @xmath0 is @xmath167 , for some @xmath168 , @xmath169 .
[ lm : hasinvolution ] let @xmath90 , @xmath170 , @xmath171 , @xmath172 .
then @xmath173 contains an involution .
we may assume that @xmath174 , @xmath175 , for some @xmath152 , @xmath153 , @xmath176 , @xmath177 .
then exactly one of the two elements @xmath178 , @xmath179 is of order @xmath25 .
this means that @xmath180 does not divide the order of any subgroup of @xmath0 .
every group of order @xmath181 contains an element of order at least @xmath182 ( the only two nonabelian groups of order @xmath181 with sylow @xmath25-subgroups isomorphic to @xmath183 are @xmath184 and @xmath185 ) .
hence @xmath186 , @xmath25 , @xmath142 , @xmath182 , @xmath187 , @xmath188 , @xmath189 , @xmath190 , @xmath191 , @xmath192 for every subgroup @xmath107 of @xmath0 .
it is not obvious , at least to the author , that @xmath0 contains no subgroups of order @xmath190 , necessarily isomorphic to @xmath193 .
it is true , however , and we prove it in section [ sc : lattice ] .
hence @xmath194 , @xmath25 , @xmath142 , @xmath182 , @xmath187 , @xmath188 , @xmath195 .
chein concludes in ( * ? ? ?
xii ) that every small moufang loop containing no element of order greater that @xmath142 is necessarily of the form @xmath111 for some nonabelian group @xmath107 .
thanks to the restrictions on @xmath196 , there are only two candidates for @xmath107 , namely @xmath197 ( the symmetric group of order @xmath187 ) and @xmath198 ( the alternating group of order @xmath189 ) .
[ cr : possiblesubloops ] a nontrivial subloop of @xmath0 is isomorphic to @xmath199 in particular , @xmath0 has the strong lagrange property .
all loops listed in @xmath200 actually occur as subloops of @xmath0 , as we shall see .
we construct three kinds of automorphisms of @xmath0 . [
lm : lieauto ] let @xmath201 be a nonsingular linear transformation . define @xmath202 by @xmath203 then @xmath204 if @xmath71and only if@xmath205 @xmath206 is an automorphism of the lie algebra @xmath207 .
linearity is obvious and the rest follows by straightforward computation using zorn s multiplication @xmath208 .
identify @xmath209 with the linear transformation @xmath210 , @xmath211 , @xmath212 , @xmath213 , @xmath214 , @xmath215 . by lemma [ lm : lieauto ] , @xmath216 . to keep the terminology simple
, we will call such automorphisms permutations ( of coordinates ) .
define @xmath217 by @xmath218 and verify that @xmath219 .
finally , we focus on conjugations .
not every conjugation in a moufang loop is an automorphism .
[ lm : conj ] let @xmath220 be a simple moufang loop . for @xmath221 define @xmath222 by @xmath223 .
then @xmath224 is a nontrivial automorphism of @xmath220 if and only if @xmath147 . by
* thm iv.1.6 ) , @xmath224 is a pseudo - automorphism with companion @xmath225 .
so @xmath224 is an automorphism whenever @xmath226 divides @xmath142 .
conversely , if @xmath224 is a nontrivial automorphism of @xmath220 , then it must be a pseudo - automorphism with companions @xmath227 and @xmath225 . by (
* thm iv.1.8 ) , the set of all companions of @xmath224 equals @xmath228 , where @xmath229 is the nucleus of @xmath220 . since @xmath220 is simple , we must have @xmath230 .
the conclusion of lemma [ lm : lieauto ] remains valid over any finite field @xmath93 , but we then get @xmath231 , rather than @xmath232 . the map @xmath233 defined by @xmath234 is an automorphism if and only if @xmath95 is even .
the detailed discussion concerning @xmath2 starts here .
[ lm : involutions ] let @xmath149}_{n}}$ ] , @xmath236}_{m}}$ ] be two involutions , @xmath237 , and let @xmath238 be an element of order @xmath142 in @xmath0 .
then : 1 .
@xmath239 if and only if @xmath240 if and only if @xmath241 , 2 . @xmath242 if and only if @xmath243 if and only if @xmath244 , 3 .
@xmath90 is contained in a copy of @xmath197 , 4 .
every copy of @xmath197 contains an involution of the form @xmath245}_{0}}$ ] , 5 .
@xmath246 if and only if @xmath247 , 6 .
@xmath248 is contained in a copy of @xmath197 .
the involution @xmath90 commutes with @xmath249 if and only if @xmath239 . since @xmath250 parts ( i ) and ( ii ) follow .
part ( v ) is proved similarly .
let @xmath149}_{n}}$ ] . without loss of generality , assume that @xmath251 .
pick @xmath176 , @xmath252 so that @xmath253 , @xmath254 . then choose @xmath255 so that @xmath236}_{m}}\in c$ ] .
then @xmath243 , and ( iii ) is proved .
let @xmath256 , @xmath257 , and suppose that @xmath149}_{1}}$ ] , @xmath236}_{1}}\in g$ ] , @xmath237 .
then @xmath258 since @xmath242 , we have @xmath244 . in other words , @xmath259 .
then the third involution @xmath260 equals @xmath261 now , @xmath262 since @xmath263 , and we are through with ( iv ) . since @xmath264
, we can assume that the first coordinate of both @xmath265 and @xmath266 is equal to @xmath56 .
then @xmath267 , @xmath268 and @xmath269 make @xmath149}_{n}}$ ] into an involution satisfying @xmath246 , by ( v ) .
this proves ( vi ) .
let us write @xmath270 for the vector @xmath271 , @xmath212 , @xmath272 , and @xmath273 for @xmath274 ( the _ weight _ of @xmath152 ) .
introduce @xmath275}$ ] as the canonical involution of @xmath0 .
observe that when @xmath90 , @xmath249 are two involutions of @xmath0 generating a subgroup isomorphic to @xmath197 then @xmath276 maps @xmath90 to @xmath249 .
[ pr : transitiveonc2 ] @xmath1 acts transitively on the @xmath156 copies of @xmath235 in @xmath0 .
we show how to map an arbitrary @xmath149}_{n}}$ ] onto @xmath277 . by lemma [ lm : involutions](iii )
, @xmath90 is contained in a copy of @xmath197 .
then , by lemma [ lm : involutions](iv ) and the observation immediately preceding this proposition , we can assume that @xmath278 .
let @xmath279 , @xmath280 .
using @xmath281 from section [ sc : autos ] , we can assume that @xmath282 .
we now fix @xmath283}$ ] and proceed to transform @xmath90 into @xmath284 so that @xmath285 , or @xmath286 , or @xmath287 , or @xmath288 .
when @xmath289 , then @xmath290 , by lemma [ lm : involutions](ii ) .
so assume that @xmath291 . since @xmath278 , we have @xmath292 , and thus @xmath293 , @xmath294 , @xmath295 , or @xmath296
. every permutation of coordinates can be made into an automorphism of @xmath0 , as we have seen in section [ sc : autos ] .
moreover , @xmath277 is invariant under all such permutations .
when @xmath293 , transform @xmath90 into @xmath249 . when @xmath297 , transform @xmath90 into @xmath298}$ ] , and note that @xmath288 .
when @xmath299 , transform @xmath90 into @xmath300}$ ] , and note again that @xmath288 . finally , when @xmath301 , we have @xmath302
now , when @xmath287 or @xmath288 , we can permute the involutions of @xmath0 so that @xmath284 is mapped to @xmath277 or @xmath249 , respectively .
since @xmath303 , we are done .
note that , in spirit of @xmath15 , the proof of proposition [ pr : transitiveonc2 ] tells us how to construct automorphisms mapping involutions of @xmath0 onto the representative @xmath277 .
in this section , we apply proposition [ pr : counting ] for the first time . we have taken advantage of the fact that all permutations and @xmath281 leave @xmath277 invariant . to proceed further , we need additional automorphisms with this property .
consider @xmath305 , @xmath306 , and define @xmath307 by @xmath308 .
then @xmath309 , by lemma [ lm : conj ] , and @xmath310 .
set @xmath311}$ ] , and let @xmath312 be the canonical element of order @xmath142 .
[ pr : transitiveonc3 ] @xmath1 acts transitively on the copies of @xmath197 and @xmath304 .
since @xmath313 and @xmath314 , by lemma [ lm : involutions](vi ) , it suffices to prove that @xmath1 acts transitively on the copies of @xmath197 .
let @xmath257 , @xmath315 , @xmath316 . by proposition [ pr : transitiveonc2 ]
, we can assume that @xmath302 .
write @xmath317}_{n}}$ ] , @xmath279 , @xmath280 . by lemma [ lm : involutions ]
, we have @xmath318 . using @xmath281
, we can assume that @xmath319 .
we are going to transform @xmath249 into @xmath320 .
assume that @xmath269 .
then @xmath262 .
taking permutations of coordinates and the possible values of @xmath321 into account , we transform @xmath249 into one of @xmath322}$ ] , @xmath323}$ ] , @xmath324}$ ] , @xmath325}$ ] . with @xmath326 as above , check that all of @xmath327 , @xmath328 , @xmath329 and @xmath330 have zeros on the diagonal .
we may hence assume that @xmath278 .
then @xmath331 , and we are done by permuting coordinates .
@xmath332 , @xmath333 , @xmath334 .
pick an involution @xmath90 . by proposition
[ pr : transitiveonc3 ] , the number of involutions @xmath249 satisfying @xmath242 is independent of @xmath90 .
one can then immediately see with @xmath335}$ ] , say , that there are @xmath191 such involutions . as @xmath336 ,
we get @xmath332 . then , by @xmath86 , @xmath337 . again by @xmath86 , @xmath338 .
note that lemma [ lm : involutions ] allows us to construct all copies of @xmath197 containing @xmath277 , and also all copies of @xmath197 containing @xmath339 .
note further that we did not have to resort to local analysis to find the value of @xmath340 . from this moment on , we will pay less attention to @xmath17 and @xmath15 .
fix @xmath341 , and recall that @xmath342 , @xmath343 . [ pr : transitiveona4 ] @xmath1 acts transitively on the @xmath156 copies of @xmath198 , and@xmath344 .
working in @xmath0 , we have @xmath345 or @xmath198 for every copy @xmath107 of @xmath304 and every involution @xmath90 . since @xmath334 and @xmath336 , there are @xmath346 involutions @xmath90 in @xmath107 such that @xmath345 . thus @xmath347 . by @xmath86 , @xmath348 . as for the transitivity , pick @xmath349 . by proposition [ pr : transitiveonc2 ]
, we can assume that @xmath350 , for some @xmath238 with @xmath351 , @xmath352 .
since @xmath353 , we have @xmath318 , by lemma [ lm : involutions](v ) , and may thus assume that @xmath354 .
then @xmath331 is the only possibility , and @xmath248 can be transformed to @xmath355 or @xmath356 .
perhaps it would be more natural to look at the copies of @xmath357 now , however , the klein subgroups of @xmath0 are exceptional in the sense that @xmath1 does not act transitively on them ( cf . section [ sc : e4 ] ) .
we therefore proceed towards @xmath142-generated subloops instead .
have a look at table [ tb : main ] .
it lists all involutions of @xmath0 and their relation to a few chosen elements of @xmath0 .
# 1#2#3#4#5#6 l#1 # 2 # 3 + # 4 # 5 # 6 @xmath358 by @xcite or @xcite , @xmath359 . by lemma [ lm : structuralm2n](i ) ,
@xmath360 .
introduce @xmath361}$ ] .
[ pr : transitiveonm12 ] @xmath362 .
in particular , @xmath1 acts transitively on the @xmath363 copies of @xmath136 . in view of proposition [ pr : transitiveonc3 ]
, it suffices to prove @xmath364 and count the copies of @xmath136 . by @xmath365 , @xmath136
is presented by @xmath366 .
using lemma [ lm : involutions ] , verify that @xmath302 , @xmath367 and @xmath368 satisfy these presenting relations , i.e. , that @xmath369 . let @xmath107 be a copy of @xmath197 . by proposition [ pr : transitiveonc3 ] , we can assume that @xmath370 . according to table [ tb : main ] ,
there are @xmath187 involutions @xmath371 such that @xmath372 ( recall that @xmath373 ) .
since @xmath374 consists solely of involutions , we have proved @xmath375 . by @xmath86 , @xmath376 .
we have @xmath377 as a special case of lemma [ lm : structuralm2n](v ) .
introduce @xmath378}$ ] .
[ pr : transitiveonma4 ] @xmath379 .
in particular , @xmath1 acts transitively on the @xmath156 copies of @xmath137 . in view of proposition [ pr : transitiveona4 ]
, it suffices to prove @xmath380 and count the copies of @xmath137 . by @xmath365 , @xmath137
is presented by @xmath381 .
verify that @xmath302 , @xmath382 and @xmath383 do the job , hence @xmath369 .
let @xmath107 be a copy of @xmath198 . by proposition [ pr : transitiveona4 ]
, we can assume that @xmath384 .
then @xmath385 is an involution . according to table [ tb : main ] , there are @xmath386 involutions @xmath371 such that @xmath387 .
( one of them is @xmath388 . )
that is why @xmath375 . by @xmath86 , @xmath389 .
let us calculate a few more hasse constants .
[ lm : morehasse ] we have @xmath390 since @xmath342 , @xmath86 yields @xmath344 .
as @xmath391 is known ( lemma [ lm : structuralm2n ] ) , the value of @xmath392 can be calculated by @xmath86 , too .
it remains to find @xmath393 and @xmath394 .
let @xmath395 , where @xmath349 .
every subgroup of @xmath396 isomorphic to @xmath197 can be written as @xmath397 for some @xmath90 , @xmath398 , @xmath147 , and there are exactly six choices of @xmath399 . since @xmath343 , we have @xmath400 . consequently , @xmath401 .
as announced before , we show that @xmath1 does not act transitively on the copies of @xmath357 .
introduce @xmath402}$ ] .
[ lm : atmosttwo ] let @xmath173 be one of the @xmath403 copies of @xmath357 in @xmath0 . then there is @xmath404 such that @xmath405 and @xmath406 . recall that @xmath332 .
therefore , given any involution @xmath90 , there are @xmath407 involutions @xmath249 such that @xmath408 .
hence , @xmath409 .
as always , we may assume that @xmath302 , @xmath317}_{n}}$ ] , @xmath410 , @xmath411 , @xmath291 , and @xmath412 .
when @xmath413 , transform @xmath249 into @xmath414 ; if @xmath293 , into @xmath415 or @xmath416 , depending on @xmath21 ; if @xmath417 , into @xmath418}$ ] ; if @xmath297 , into @xmath419}$ ] or @xmath420}$ ] .
recall the automorphism @xmath326 from section [ sc : c3 ] , and check that @xmath421 , @xmath422 , @xmath423 , @xmath424 .
thus @xmath425 can be transformed into @xmath415 , and each of @xmath414 , @xmath426 , @xmath427 can be transformed into @xmath416 .
assume , for a while , that @xmath1 acts transitively on the @xmath403 copies of @xmath357 .
then , by @xmath86 , @xmath428 , a contradiction
. hence , by lemma [ lm : atmosttwo ] , there are @xmath25 orbits of transitivity @xmath429 , @xmath430 , with representatives @xmath431 , @xmath432 . since @xmath433 , we have @xmath434
. then @xmath435 .
by @xmath85 , @xmath436 and , similarly , @xmath437 .
by @xmath73 , @xmath438 .
this is only possible when @xmath439 and @xmath440 . in other words , a copy of @xmath357
is contained in @xmath198 if and only if it belongs to @xmath429 .
let us have a look at the relation between @xmath357 and @xmath136 .
[ lm : usecosetdistribution ] @xmath441 , @xmath442 .
consider @xmath431 .
assume that there is @xmath257 such that @xmath443 . since @xmath444 , @xmath445 , @xmath446 , @xmath447 , there is exactly one involution @xmath448 in @xmath107 , say @xmath445 .
write @xmath449}$ ] for appropriate vectors @xmath450 , and note that @xmath451
. there is @xmath452}\in g$ ] such that @xmath453 .
then @xmath454 , @xmath455 .
by lemma [ lm : involutions ] , @xmath456 if and only if @xmath457 .
hence @xmath458 , a contradiction .
inevitably , @xmath441 . we proceed to calculate @xmath459 .
since @xmath460 , by lemma [ lm : structuralm2n ] , we have @xmath461 .
finally , we have a look at the constants @xmath462 , for @xmath463 . with the above notation for @xmath464 , @xmath465 , we have 1 .
@xmath466 , @xmath467 , @xmath468 , @xmath469 , @xmath470 , 2 .
@xmath471 , 3 .
@xmath472 .
hence @xmath473 and @xmath474 .
since @xmath342 and @xmath475 , we have @xmath476 , by lemma [ lm : structuralm2n ] . formula
@xmath74 then yields @xmath477 .
in particular , @xmath478 , and ( i ) follows .
let @xmath431 .
we are trying to find a group @xmath349 such that @xmath479 .
we look again at the distribution of the @xmath142 involutions @xmath277 , @xmath415 , @xmath480 in the cosets @xmath107 , @xmath114 .
there are two possibilities : either @xmath481 , or @xmath482 .
suppose that @xmath481 .
as @xmath475 and @xmath377 , there is at most one subloop @xmath483 such that @xmath484 in such a case .
now suppose that @xmath482 .
then @xmath485 is one of the three @xmath25-element subgroups of @xmath486 .
let us call it @xmath487 .
since @xmath342 and @xmath377 , there are at most @xmath142 subloops @xmath488 such that @xmath489 . because there are three ways to choose @xmath487 in @xmath486 , there are at most @xmath490 subloops @xmath483 such that @xmath484 .
altogether , @xmath491 . by ( i ) ,
@xmath471 , and ( ii ) is finished , let @xmath432 .
we are trying to find a group @xmath349 such that @xmath492 . since @xmath493 , the group @xmath494 is not contained in @xmath107 , i.e. , @xmath495 . by proposition
[ pr : transitiveonc2 ] , we can assume that @xmath496 .
if there is such a group @xmath107 , there is also an element @xmath497 such that @xmath498 , i.e. , @xmath499 by lemma [ lm : involutions ] , the system of equations @xmath500 is equivalent to @xmath501 in particular , @xmath502 .
there are @xmath182 solutions to @xmath503 , namely @xmath504 this is easy to see since both @xmath505 , @xmath506 lead to @xmath507 .
hence , there are at most @xmath188 candidates for @xmath249 ( with @xmath278 , @xmath56 ) .
however , if @xmath508 is isomorphic to @xmath198 , then every element of order @xmath142 in @xmath508 must satisfy @xmath503 .
there are @xmath188 elements of order @xmath142 in @xmath198 , and thus there is at most @xmath56 subloop @xmath509 satisfying all of our restrictions . because our choice of @xmath510 was one of three possible choices ,
we conclude that @xmath472 .
combine ( i ) , ( ii ) , ( iii ) to get @xmath473 , @xmath474 .
recall the representatives @xmath431 , @xmath432 , and observe that the loop @xmath511 is a group isomorphic to @xmath183 .
[ lm : dminusdplus ] @xmath512 , @xmath513 .
write @xmath514 .
we have seen that both @xmath515 , @xmath516 are positive .
inspection of table [ tb : main ] reveals that there are @xmath189 involutions @xmath517 such that @xmath518 .
this immediately shows that @xmath519 .
in fact , @xmath520}$ ] , @xmath521}$ ] , @xmath522}$ ] yield @xmath142 different copies of @xmath183 . thus @xmath523 .
yet another inspection of table [ tb : main ] shows that there are @xmath189 involutions @xmath524 such that @xmath525 .
this means that @xmath526 , but we prove more . the group @xmath494 is contained in @xmath182 copies of @xmath136 , by lemma [ lm : usecosetdistribution ] .
let @xmath396 be one of them .
we can assume that @xmath527 , where @xmath257 , @xmath528 , @xmath529 .
since @xmath530 , every element @xmath249 of @xmath114 satisfies @xmath531 .
no involution @xmath249 of @xmath107 satisfies @xmath531 . since @xmath532 and @xmath533 , there are @xmath142 involutions @xmath534 such that @xmath535
. one of them is @xmath536 .
altogether , @xmath537 .
[ lm : insertedlater ] every copy of @xmath183 in @xmath0 contains a subgroup from @xmath430 .
note that the proof of lemma [ lm : atmosttwo ] implies that @xmath538 if and only if @xmath249 is a permutation of @xmath415 or @xmath425 , i.e. , @xmath249 is one of the @xmath187 diagonal elements in table [ tb : main ] .
let us denote this set by @xmath27 .
let @xmath539 be a copy of @xmath183 in @xmath0 . without loss of generality , @xmath540 .
assume that @xmath538 for every @xmath541 .
then @xmath542 .
we proceed to show that @xmath543}$ ] , @xmath283}\in s$ ] satisfy @xmath544 .
we have carefully chosen the notation so that @xmath249 is the same as in the proof of proposition [ pr : transitiveonc2 ] . according to the last line of that proof , @xmath545 . using the same automorphism again
, we get @xmath546}$ ] .
hence @xmath544 .
[ pr : transitivee8 ] the group @xmath1 acts transitively on the @xmath156 copies of @xmath183 .
also , @xmath547 .
let @xmath539 , @xmath548 be two subgroups of @xmath0 isomorphic to @xmath183 .
then there are @xmath107 , @xmath549 such that @xmath550 , @xmath551 , by lemma [ lm : insertedlater ] . since @xmath107
, @xmath552 belong to the same orbit , there is @xmath404 mapping @xmath107 onto @xmath552 . as @xmath513 , @xmath266 must map @xmath539 onto @xmath548 . by @xmath74 and lemma [ lm : dminusdplus ] , @xmath553
hence , @xmath554 .
consequently , @xmath86 yields @xmath555 , and we are finished .
it is about time to show that @xmath0 contains no copies of @xmath193 .
assume that @xmath556 is a subgroup of @xmath0 .
by proposition [ pr : transitivee8 ] , we can assume that @xmath557 .
then there must be at least @xmath188 involutions @xmath249 outside @xmath511 in @xmath0 such that @xmath558 .
previous inspection of table [ tb : main ] provided none , a contradiction .
let us summarize the results about @xmath18 obtained in this paper .
[ th : main ] the smallest @xmath141-element nonassociative simple moufang loop @xmath0 satisfies the strong lagrange property but not the weak cauchy property .
the following loops @xmath71and no other@xmath205 appear as subloops of @xmath0 : @xmath559 , @xmath235 , @xmath304 , @xmath357 , @xmath197 , @xmath183 , @xmath198 , @xmath136 , @xmath137 , and @xmath0 .
the automorphism group @xmath1 acts transitively on the copies of each of these subloops , with the exception of @xmath357 .
there are two orbits of transitivity @xmath429 , @xmath430 for @xmath357 . with the notational conventions introduced in section @xmath560
, we have the following orbit representatives : @xmath561 for @xmath235 , @xmath562 for @xmath304 , @xmath563 for @xmath429 , @xmath564 for @xmath430 , @xmath565 for @xmath197 , @xmath511 for @xmath183 , @xmath566 for @xmath198 , @xmath567 for @xmath136 , and @xmath568 for @xmath137 , where @xmath275}$ ] , @xmath569 , @xmath341 , @xmath361}$ ] , @xmath378}$ ] , and @xmath402}$ ] .
the subloop structure and hasse constants for @xmath0 are summarized in figure @xmath570 .
this paper is based on ( * ? ? ? * ch .
shortly after i finished working on @xcite , orin chein brought my attention to the work of merlini guiliani and polcino milies @xcite . in @xcite
, the authors studied the subloop lattice of @xmath571 for the first time , and 1 .
determined @xmath572 , @xmath573 , and estimated @xmath574 , 2 . used a result analogous to lemma [ lm : involutions](i ) , ( ii ) , 3 . found all possible isomorphism types of subloops of @xmath0 and listed one example for each type , thus establishing the strong lagrange property for @xmath0 , 4 . sketched the relations between isomorphism types of subloops of @xmath0 , m. l. merlini guiliani , csar polcino milies , _ on the structure of the simple moufang loop @xmath577 _ , in nonassociative algebra and its applications , the fourth international conference , lecture notes in pure and applied mathematics * 211 * , edited by roberto costa , alexander grishkov , henrique guzzo , jr .
, luiz a. peresi .
marcel dekker , new york ( 2000 ) . | hasse constants and their basic properties are introduced to facilitate the connection between the lattice of subalgebras of an algebra @xmath0 and the natural action of the automorphism group @xmath1 on @xmath0 .
these constants are then used to describe the lattice of subloops of the smallest nonassociative simple moufang loop . |
much of the underlying physics of a core collapse supernova remains unexplored , the mechanism leading to the explosion not well understood . though it is clear that neutrinos play a major role in the dynamics of the explosion , carrying away 99% of the energy emitted by the dying star .
the observation of neutrinos in the provenance of sn 1987a in several experiments @xcite has confirmed the crucial role they play in the overall picture .
the limited number of detected neutrinos , however , has not provided enough statistics to constrain current supernova models .
this may only be achieved with a detector performing routine sn detection , about one per year .
the idea of using icecube as a supernova detector was first discussed in @xcite .
the method is based on the sudden increase in the photomultiplier count rate on a timescale of the order of @xmath3seconds : qualitatively , the release by a supernova at @xmath4kpc with a total amount of energy of @xmath5mev in neutrinos of about @xmath6mev corresponds to a flux of @xmath7 in its completed configuration and with its data acquisition system , icecube has about @xmath8mton effective detection volume and a reach of about @xmath9kpc at @xmath10 c.l .
detection level . here
, we extend that method and assess the potential of icecube and several prospective detector configurations by considering multiple hit detection in a single or neighboring sensors from individual neutrino interactions .
we demonstrate that multi - hit modes enable the extraction of new observables related to the features of the neutrino emission spectrum and the supernova location , not accessible by means of the original method . this new data stream could help study questions related to supernova dynamics ( neutronization burst , accretion and cooling phase , see @xcite and references therein ) or fundamental questions of particle physics , related to the nature of the neutrino @xcite .
the observation of electron neutrinos from the early neutronization burst could disentangle the neutrino hierarchy @xcite , provided a sufficiently large mixing angle @xmath11 as indicated recently @xcite .
these new detection modes are conceptually situated between a full reconstruction of the events and the current search method balancing the simplicity of the latter with the richness in information of the former .
we show that the loss of information in the current method is not necessary and that a very elementary reconstruction , implemented through the coincident hit methodology , can enhance the information obtained during a supernova explosion .
in fact , it would appear natural to extend the new detection modes to a full reconstruction of supernova events in the case of a very densely instrumented volume . according to @xcite ,
the estimated sn rate from our galaxy is about @xmath12 .
it increases to @xmath13 within @xmath14mpc and about @xmath15 within @xmath16mpc .
we will see that extending the reach of a future detector to such a large distance is quite challenging given the quick decrease of the interaction density with distance .
superk , with its @xmath17kton fiducial volume , remains at the moment an instrument of choice @xcite .
icecube @xcite with its 86 strings and this new detection methodology could become a serious competitor , even surpass it , provided a deep and denser core extension installation .
there are several good reasons for such an extension beyond an enhanced potential for sn detection , mainly the detection of individual high energy neutrinos down to several gev related to multiple physics aspects , such as neutrino sources in the southern sky @xcite , indirect detection of dark matter down to lower mass @xcite and neutrino particle physics @xcite .
this contrasts with the potential of other projected @xmath18kton - class detectors , e.g. liquid scintillator lena @xcite or liquid argon detectors @xcite , which will have an extended potential for sn detection but comparatively small fiducial volumes at energies beyond gev .
especially the latter are complementary to water or ice - based cherenkov detectors , since they are more sensitive to the electron neutrino channel , providing a better access to the physics during the neutronization burst @xcite . combining the measurements from these detectors , which differ not only in their sensitive material , but in their geographic location , would enable the study of matter effects and neutrino hierarchy by observing neutrinos from the same supernova from different viewpoints with a high detection probability .
finally , a large detector will not only increase the distance reach for sn but also provide large event statistics for galactic sn with possible resolution for their location , leading through a sn alert system ( snews @xcite ) to the optical detection of obscured and otherwise missed galactic sne . in section 2 , we define the studied configuration and the benchmark supernova neutrino flux .
we present the effective volume analytical calculations in section 3 and their consequences in section 4 .
in the following , we consider several detector configurations @xcite : [ [ the - baseline - icecube - configuration ] ] the baseline icecube configuration + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + ( ic ) , with a number of sensors @xmath19 , arranged @xmath20 m apart along 80 strings .
[ [ the - deep - core - configuration ] ] the deep core configuration + + + + + + + + + + + + + + + + + + + + + + + + + + + ( dc ) , with @xmath21 , @xmath22 m apart on 8 strings .
[ [ the - dense - and - deep - core - configuration ] ] the dense and deep core configuration : + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + an imaginative futuristic low energy extension in alternative configurations denoted ddc , ddc@xmath23 , ddc@xmath24 composed of 24 strings with 150 sensors each , @xmath25 , and with a spacing between modules of @xmath26 m . in configurations with @xmath27 subscript , the sensitive area of the spherical module has a @xmath28 acceptance , for instance the modules designed for km3net @xcite .
the ddc@xmath24 configuration has also a larger sensitive area and higher q.e .
in addition , we assess the potential of the full acceptance configurations with extremely low noise modules ( @xmath18hz ) .
a short local coincidence is the occurrence of two or more hits within specific time windows @xmath29 s in a single sensor or in two neighboring sensors , also referred to as optical modules .
the systematic recording of these coincidences would require the implementation of a dedicated data acquisition mode and its corresponding data stream .
we study in this paper several detection modes @xmath30 , @xmath31 and @xmath32 , referring to the coincidental detection of @xmath33 hits with a module and @xmath34 hits with the module just above along a sensor string . in this paper
we do not consider coincident hits between strings , so string spacing is not a parameter in this study .
we refer therefore to our study as 1d analytic approximation .
we also restrict the study to nearest neighbor ( nn ) coincidences .
the geometry of the optical modules for the three detector configurations ic , dc and ddc is illustrated in fig .
[ fig : schematic ] .
the modules @xcite are composed of a @xmath35inch enveloping glass sphere ( to withstand the pressures at the large deployment depths , in particular during the ice refreezing process ) with an external radius @xmath36inch , enclosing a photomultiplier r7081 - 02 made by hamamatsu photonics @xcite .
the glass and contact gel between the pmt and the glass introduce a @xmath37 nm lower cutoff . in all configurations except ddc@xmath23
, the photocathode sensitive surface of the sensor covers a fraction of the sphere slightly smaller than the half sphere .
the resulting cutoff angle on the sensitive surface is given by @xmath38 , where @xmath39 is the limit on the incoming direction of cherenkov photons at which they can intersect the sensitive surface of the module .
given the specifications of the icecube photomultiplier , we can estimate @xmath40 .
viewed under an angle @xmath41 , the projected area of the sensitive surface is @xmath42 where @xmath43 for a module sphere looking down and with a photocathode sensitive area cutoff angle @xmath38 , as illustrated on fig . [
fig : schematic ] .
the photon detection efficiency ( pde ) of the sensitive surface @xmath44 has been measured somewhere @xcite , it combines the pmt q.e . , the transparency of the surrounding glass and the contact gel which couples the glass to the pmt . averaged over a @xmath45 cherenkov spectrum between @xmath46 and @xmath47 nm , and accounting for an amplitude trigger threshold keeping @xmath4890% of the p.e.s , it yields an integrated pde @xmath49 for the regular ic sensors .
the high quantum efficiency sensors deployed in dc are assumed to have a pde 30% higher @xcite . the same assumption is made for the ddc and ddc@xmath23 configurations .
in contrast to previous works @xcite , @xmath50 is significantly lower , due to a regular optical module peak pde of about 14% at @xmath51 nm .
although @xmath50 is a fair approximation for our analytical estimates presented in the next section , a toy monte carlo simulation should implement the complete wavelength dependence of @xmath44 given that the propagation of photons in the ice medium is described in terms of wavelength - dependent scattering and absorption lengths @xmath52 and @xmath53 @xcite .
this dependence in the propagation can not be disentangled from that of the pde .
the following types of background may limit the applicability of the coincident hit detection methodology : [ [ cosmic - ray - induced - muons . ] ] cosmic - ray induced muons .
+ + + + + + + + + + + + + + + + + + + + + + + + + a fraction will travel down to the detector and produce coincident hits .
we assume that most of these events are muons dying in the upper layers of the detector : they barely enter the fiducial volume and produce light on a very limited number of external sensors .
we assume that this event class can be adequately rejected .
some higher energy muons however penetrate deeper into the detector and although most of them can be vetoed given their specific topology , some may leave a signature well inside the instrumented volume mimicking the coincidence signature under investigation .
we would expect the resulting rate to be minimal and thus neglect it , but this specific topology should still be further studied separately by means of the experimental data , to properly substantiate this statement .
[ [ low - e - atmospheric - neutrino - induced - muons - or - electronic - showers . ] ] low @xmath54 atmospheric neutrino - induced muons or electronic showers .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + this background is constituted by low energy neutrinos in the @xmath55 range @xcite .
the interaction rate is very limited , well below @xmath14hz , in a gton ice volume .
[ [ sensor - correlated - noises - at - mus - and - ms - time - scales - due - respectively - to - photomultiplier - afterpulse - and - radioactive - decay - chains ] ] sensor correlated noises at @xmath56s and ms time scales , due respectively to photomultiplier afterpulse and radioactive decay chains + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + @xcite .
the afterpulse components of the dark noise come with well defined @xcite delays after the original photo - induced signal , larger than the time windows used below for coincidental hit detection within a single sensor , so that ( 1 ) it does not pollute and therefore impede the coincidental hit detection method for this specific channel , ( 2 ) it can be filtered out by means of an offline analysis , leading to an effective noise rate significantly smaller than measured ( dark count reduced from @xmath57 hz to @xmath58hz @xcite ) .
this reduced dark count rate value is thus considered to calculate the probability of the noise .
note that the deadtime introduced by this afterpulse filter is negligible ( @xmath59 0.5% ) .
the effect of ms timescales of correlated noise is more complex and it is not clear whether it would affect the potential of the new methodology when applied to specific channels with multiple hits in a single sensor .
we define the hit times @xmath60 , a hit time window by its width @xmath61 and an arbitrary time shift @xmath62 .
hits @xmath33 and @xmath34 belong to the time window if @xmath63 .
we restrict the study to several qualitatively complementary nn detection modes , * single sensor hit : mode @xmath64 or `` 1 + 0 '' ; * two photons detected in the same sensor : mode @xmath65 or `` 2 + 0 '' ; * a single hit detected in two nn sensor : mode @xmath66 or `` 1 + 1 '' ; * three photons detected at times @xmath67 ( in the same sensor , @xmath68 ) and @xmath69 ( in the sensor above ) : mode @xmath70 or `` 2 + 1 '' .
the noise rates of two hit coincidences @xmath71 and @xmath72 given by @xmath73 ( the latter factor two because hits at a single module are indistinguishable ) .
the coincidental noise rate with two hits in the lower sensor and a single hit in its upper neighbor is @xmath74 note that in the icecube detector , only the lower hemisphere of the optical module is sensitive and therefore , a hit in the upper sensor coming after a hit in the lower sensor is strongly preferred ( i.e. an optimal time window will have @xmath75 ) .
this is not true for some of the futuristic detector configurations which have a uniform sensitive surface : in this case , we will choose @xmath76 .
the detector coincidental noise rate scales linearly with the number of sensors ( the number of independent pairs is almost equivalent to the number of sensors , provided that the number of sensor along a string is @xmath77 ) . icecube has a system that records local coincidences within a @xmath78s ( @xmath79s ) time window up to the next - to - nearest neighbor ( there are four such neighbors ) @xcite .
the rate of such coincidences , called _ hard local coincidences _ ( hlc ) , should be therefore of the order of @xmath80hz per optical module . for our purpose
, we consider the rate of coincidental events within a much shorter coincidence time window with @xmath81ns involving one neighboring sensor corresponding to a coincidental noise rate of @xmath82hz per sensor .
a sn signal becomes significant when the number of coincidental events integrated over the typical duration of the sn neutrino emission ( about 3 seconds ) takes values larger than the fluctuations of the noise rate @xmath83 , corresponding to the various event classes .
we have chosen to study several flux models of rather modest luminosity .
our main benchmark flux model is the `` sf '' model ( also referred to as `` garching '' model ) , an 8.8 solar mass electron capture supernova , discussed in @xcite and which may represent a large fraction of all supernovae .
furthermore , in this paper we will only consider the predicted anti - electron neutrino flux , focusing solely on the inverse beta decay ( ibd ) reaction in ice , _ i.e. _ @xmath84 .
for one , in water or ice , which are natural media for high energy neutrino telescopes , the ( anti-)electron neutrino interaction with oxygen has a much higher energy threshold and a quite significantly smaller cc cross section at the energies of interest @xcite .
this largely makes this channel subdominant for the interaction rate in the detector compared to the ibd reaction yield .
moreover , a large fraction of the neutrino energy is absorbed into binding energy , reducing the cherenkov light yield and dramatically suppressing the potential of this reaction channel given that our methodology relies on the coincident detection of cherenkov photons in neighboring sensors .
we also neglect the elastic interaction channel @xmath85 which has an even lower cross section than the oxygen channel @xcite .
the anti - electron neutrino flux luminosity is given by @xmath86 where @xmath87 and with the time - dependent model parameters luminosity @xmath88 , average neutrino energy @xmath89 and pinch parameter @xmath90 .
their time evolution can be found in @xcite .
it is interesting to note , as illustrated in fig .
[ fig : timeintpositronspectrum ] , that the parameters @xmath91 and @xmath92 can be very well fitted with eq .
[ eq : falphaavenu ] in order to match the integrated neutrino flux and positron density up to a certain time . .
dashed line is the result of fitting @xmath93 and @xmath94mev.[fig : timeintpositronspectrum],scaledwidth=50.0% ] the neutrino flux integration time @xmath95 , which closely optimizes the potential in terms of distance reach of the coincident hit method is derived by maximizing the signal to the square root of noise ratio , assuming an exponential luminosity decrease , in good qualitative agreement @xmath96s , with @xmath97s , of the luminosity evolution in the garching model ( a usual luminosity decrease in @xmath98 does not fit well ) .
the argument could be more sophisticated , e.g. by using a refined likelihood temporal analysis . the integrated signal ( the supernova neutrinos producing an event ) up to time @xmath99 is @xmath100 .
the detector noise is assumed gaussian with count fluctuation of a complete detector configuration @xmath101 within an integration time @xmath95 , where @xmath102 is the noise rate of a specific mode .
@xmath95 is determined @xmath103 leading to @xmath104 we extract @xmath105s for @xmath106s and we note that @xmath95 is independent of @xmath107 ( proportional to the luminosity ) , of the noise rate @xmath102 and of the significance of the observation .
the luminosity at early times is larger than in this description , so that this integration time @xmath99 constitutes a conservative upper bound regarding the potential of the method : if more precisely assessed , @xmath95 and thus the integrated noise would be reduced . the integrated anti - electron neutrino luminosity within this time - span @xmath95 is @xmath108erg , and the flux @xmath109 .
the average neutrino energy is @xmath110mev .
@xmath111 is about three quarter of the total luminosity .
it is important to note , when comparing the limiting distance up to which a sn can be discovered , that the total luminosity is about 56% of the one considered in @xcite .
we have also calculated the outcome of the novel detection methodology for neutrino spectra which obey a fermi - dirac distribution , @xmath112 where @xmath113 .
we have normalized the amount of energy released over @xmath95 to be equal to the sf model , for two temperature , @xmath114mev and @xmath115mev .
although the average neutrino energies , @xmath116mev and @xmath117mev , are larger resulting in smaller flux than in the sf model , the interaction density is increased by more than 50% due the ibd cross section increase turning into a considerable increase in the coincidental hit rate ( by a factor @xmath118 ) as will be seen .
we assess the discovery and learning potential by requesting the signal over square root of noise ratio @xmath119 and @xmath120 for the c.l . in @xmath121 unit , respectively , using the poissonian distribution @xmath122 : given a mean number of expected coincidental noise events @xmath123 , a minimum number of @xmath124 events is required to reach significance @xmath125 , @xmath126 the distance reach is therefore given by @xmath127 where @xmath128 is the distance in 10 kpc units and @xmath129 is the number of signal events from a sn explosion at 10 kpc .
an advantage of the detection method relying on coincidental hits arises from a more gaussian distributed coincidental noise rate ( while the single hit noise rate distribution has non gaussian tails , requesting coincidences dilutes them ) . the false alarm rate ( assuming gaussianity ) for @xmath119 is roughly @xmath130 .
@xmath120 is a requirement too weak and not adequate for icecube to trigger as a standalone detector , but where icecube may help establishing stronger constraints by enhancing the quality of an observation : it sets the distance scale for icecube to contribute in supplementing information to observations made by other instruments detecting e.g. a neutrino or an optical signal . in the latter case , of course
, icecube would have to buffer the relevant information because of the few hours delay between the neutrino and the optical signal .
the positron effective detection volume as a function of the neutrino energy is the product of the number of target protons in the interaction volume and of the fraction of uniformly distributed positrons leading to a detection .
it is an important and well defined concept for our analysis because the positron tracks are nearly point - like with a path length of about 10 cm in average , a small distance compared to the mean distance between a positron and a sensor in which it leaves a signature .
this enables the estimation of the response of the detector to a given flux of low energy neutrinos .
the positron effective volume for coincidence of order @xmath131 at a single module is denoted @xmath132 , where @xmath133 for the detection of a single hit resp .
of two coincident hits .
we write @xmath134 for a detection of order @xmath131 in one sensor and of order @xmath135 at the nn .
the effective volume depends only slightly on the incoming neutrino direction @xmath136 , due to the proximity of the average neutrino energy with the energy at which the anisotropy in the positron emission ( transition between backward and forward scattering ) from the ibd interaction vanishes .
note however that selecting events in the high energy tail of the neutrino emission , where forward scattering is favored , may grant access to the direction of a nearby supernova .
this can be done to some extent by means of the coincident hit time analysis .
the current supernova detection strategy with icecube relies only on @xmath137 .
we analyze here the power of the new strategy by relying more generally on @xmath134 .
we emphasize as well the importance of certain ratios , such as @xmath138 or @xmath139 , which open new perspectives and grant access to new observables , e.g. the average positron energy . for a given supernova neutrino emission model @xmath140 , time - integrated up to time @xmath95
, the effective volume definition can be extended to embed the spectral shape of the positron energy distribution @xmath141 , @xmath142 where @xmath143 and which follows from the definition @xmath144 from now on , we drop the @xmath140 or replace it by one of the specific models .
the presence of the tilde denoting time - integration necessarily refers to a model and confusion in the notation is avoided .
the differential positron density @xmath145 is extracted from the incoming neutrino flux @xmath146 by folding it with the differential ibd cross section @xmath147 given by eq . (
10 ) in @xcite ( accurate up to nnlo in @xmath148 ) , @xmath149 and the total number density of positron interactions ( same luminosity for all three models ) , @xmath150 the proton target density is given by @xmath151 , where @xmath152 , @xmath153 , and @xmath154 is the avogadro constant . the number of expected events belonging to the various classes @xmath155 ( see eq .
[ eq : nvr ] ) , obtained by integration of @xmath156 also defines the relevant observables @xmath157 , some of them being particularly sensitive to the average energy of the neutrino spectrum . in order to proceed with the effective volume estimates , we need to approximate the average number of detectable cherenkov photons emitted along the track of a positron with energy @xmath54 , which could naively be done by combining the franck - tamm @xcite ( @xmath158 ) , the average photomultiplier q.e . and positron range formulas , @xmath159 where the integrals for average q.e and franck - tamm formula run between 300 and 650 nm .
@xmath160 is the positron track length above cherenkov emission threshold and given by @xcite : @xmath161 where @xmath162 , @xmath163 and @xmath164mev @xcite .
e.g. at @xmath165mev , @xmath166 cm and the average number of detectable photons ( ic ) by a positron is @xmath167 .
however , eq . [ eq : ne ] does not take into account cherenkov photons which are emitted by secondary particles , e.g .
knock - off delta electron and positron annihilation .
results from a geant4 simulation show that the total number of emitted photons remains almost constant ( c.f .
[ fig : rangeandne - positron ] ) . in the following
we will use the geant4 photon yield results for the calculations .
the fraction @xmath168 of the @xmath169sr solid angle occupied by the sensitive surface of the module , viewed under angle @xmath41 and at distance @xmath102 is @xmath170 ( we have neglected the spherical shape of the sensor but this remains an excellent approximation ) .
when @xmath171 , this expression reduces to @xmath172 the expectation number of detected photons is @xmath173 we have now all the necessary ingredients to express the coincident hit effective volumes , @xmath174 @xmath175 is the density of probability of recording @xmath176 hits from a positron with energy @xmath177 , where @xmath178 is a poisson distribution with mean value @xmath179 for the effective volume for a sensor with @xmath131 hits and nn sensor with @xmath135 hits , we can write the more generic expression , @xmath180 where @xmath181 and @xmath182 . according to @xcite and for the purpose of this estimate , we have made the qualitative assumption that the cherenkov photons are emitted in correlated directions , roughly over @xmath183sr , or a quarter of the sphere , the approximate size of the cherenkov cone .
we therefore effectively implement the anisotropic emission with the inclusion of several factors @xmath184 . in order to obtain simple expressions ,
we have neglected the spacing between the modules which do affect the probability of two nn sensors to be in the pool of photons .
this may have a significant impact on the effective volume .
we qualitatively argue that most interactions leading to a coincidental hit detection occur slightly below the lower sensor , i.e. @xmath41 is in general close to @xmath185 .
however , this argument does not hold for 4@xmath186 sensitive modules . in this case , the expressions derived here are quite approximative .
the correlated probability of photon detection is justified at short distances from the sensor , as the photons are not yet diffused .
this correlation gradually vanishes at larger distances and fully disappears at an effective scattering length distance .
fortunately , our assumption is likely quite accurate for nn coincidental hit detection from low energy events , because the detection probability of @xmath187 photon is sizable only at short distances and the photon directions are not yet decorrelated ( i.e. the fraction of contributed events beyond a relatively short distance is negligible , due to a negligible fraction of the subtended areas of the sensors ) . the validity of the effective volume formula for single hits with the introduced anisotropic emission holds for arbitrary distance : although large values of the integration variable @xmath102 ( i.e at distances where the cherenkov photons and the initial positron directions are decorrelated ) contribute significantly to the integral , there is an equivalence at large distances between isotropic emission - uniform low photon expectation value and anisotropic emission - higher photon expectation value within the @xmath183sr cone . stated differently , the equivalence at large distances ( i.e. @xmath56 small ) @xmath188 ( see the expression for @xmath189 , eq .
[ eq : fmu ] ) . in eq .
[ eq : mur ] , we have introduced a damping factor related to the absorption length .
one may question whether we should consider instead the attenuation length as appropriate .
this can be understood from first principles ( phase space conservation ) : photons do not propagate to such large distances as they scatter and will only reach an average distance of about an attenuation length from the sensor .
however , the subtended sensor area for the photon detection decreases with distance and the large absorption length considered compensates for the shorter effective distance reached by an average photon but seeing a larger subtended photo - sensor area .
the results for the effective volume in the various detection modes are shown in fig .
[ fig : effvol ] .
it is noticeable that while the ratio of the effective volume to the positron energy for single hit detection mode behaves rather flat and even eventually decreases at high energy , it continually increases for coincidental hit detection modes .
the rate of the increase is fastest for the @xmath190 mode and faster for @xmath191 than @xmath192 .
this behavior reflects in a corresponding increase of the average positron energy for the chosen neutrino emission model .
for the four different topologies @xmath193 , the relative variation rate with energy is constant , being independent of the specific detector configuration , the absolute scaling of the effective volume , which in turn depends on the quantum efficiency , and the sensor coverage and spacing .
the effective detection volumes in the ddc configuration are significantly higher when the sensor is sensitive over 4@xmath186 , a factor about three for @xmath194 mode , four for @xmath192 mode and six for @xmath191 and @xmath190 modes around the garching energy @xmath195 . ,
ddc , dc and ic .
middle plot , thick / thin lines for resp .
@xmath192 and @xmath191 modes ( note that dc and ddc results are the same for @xmath194 and @xmath192 modes ) .
[ fig : effvol],title="fig:",scaledwidth=45.0% ] , ddc , dc and ic . middle plot , thick / thin lines for resp .
@xmath192 and @xmath191 modes ( note that dc and ddc results are the same for @xmath194 and @xmath192 modes ) .
[ fig : effvol],title="fig:",scaledwidth=45.0% ] , ddc , dc and ic . middle plot , thick / thin lines for resp .
@xmath192 and @xmath191 modes ( note that dc and ddc results are the same for @xmath194 and @xmath192 modes ) .
[ fig : effvol],title="fig:",scaledwidth=45.0% ] fig .
[ rhocoinc - vs - znu - vs - rxy ] shows the detection probability w.r.t . the ibd neutrino interaction location for dc and ddc@xmath23 configurations .
it is clear from these formulas that , while we constrain ourself to a limited number of detection modes , it is trivial to extend the analysis in order to obtain effective volumes for arbitrary coincidental modes @xmath196 , including those beyond our one dimensional analysis ( i.e. modes with hits on different strings which become relevant for a hypothetical configuration with small string spacing ) .
with the effective volumes estimated above , we now have all the ingredients to calculate the number of single and coincidental hit detections . in table [ table - sf ] , we provide the number of detections and the distance reach for unit signal over background information and for a 5@xmath121 c.l
. detection , calculated for the various configurations with the sf model , and a comparison of effective volumes in table [ table - spectra ] .
note that in the case of the fermi - dirac spectrum with @xmath115mev and the luminosity @xmath197 calculated by @xcite , and large area low noise sensors , the @xmath190 detection mode reaches @xmath198mpc and @xmath199mpc .
.sensor effective volume for the garching spectrum , @xmath200mev , various detector configurations and detection modes . the last entries , corresponding to ddc@xmath24 , are results for an enhanced ddc configuration with larger modules and q.e . , for a @xmath114mev fermi - dirac spectrum and a luminosity @xmath201 normalized to @xmath202erg.[table - sf ] [ cols="^,^,<,<,<,<",options="header " , ] it is also clear that the time difference distribution of detected coincidental hits , an observable independent from @xmath203 , should also depend on the angle of provenance of the neutrino flux @xmath136 .
a careful assessment and quantification of the resolution by studying the impact of the incoming direction on the induced distortion on the hit time difference distribution and the various coincidental hit ratios requires monte carlo studies .
such a study should also take into account the elastic scattering channel .
although we pointed out above that the low cross section of this channel will limit its contribution to the overall detection potential , the strong angular correlation and forward peaked cross section might result in a slight enhancement for the measurement of the incidence direction .
in order to confirm the simple analytical calculations for single hit effective detection volume and to precisely study the new coincident hit detection methodology and the new observables to which it grants access , precise simulation should be performed . in particular , a precise assessment of the angular resolution in locating the sn from the asymmetric cherenkov emission can be reliably studied as a function of distance and which is only qualitatively stated in this paper .
with such a tool , a dedicated work not focused only on ibd but implementing other neutrino interactions , e.g. with oxygen , would permit the precise assessment of the potential of the novel analysis method for oscillations and matter effect , hierarchy and deleptonization neutrino emission phase studies .
a significant increase in potential would be reached with the imagined ddc@xmath24 icecube extension , with sensors combining several improvements such as a larger and uniform collection surface and an increased photo - detection efficiency ( which is of specific interest as the distance reach scales linearly at first order with the q.e . for @xmath191 and @xmath192 detection modes ) .
sn detection up to @xmath204mpc would be at hand with the foreseen potential boost of this detector .
such an extension , considerably cost effective compared to similar projects , would be motivated by a broader experimental interest , vastly enhancing the detection potential of low energy neutrinos , @xmath205gev , connected to physics of neutrino oscillations , dark matter neutrino signatures , astrophysical neutrinos in particular from the southern hemisphere , etc @xcite . to reach about @xmath204mton effective detection volume , thus granting access to sn at distance scales larger of about @xmath16mpc and guarantee routine sn detection
, the ddc@xmath24 configuration is likely too modest and the number of modules should rather be about 10000 . also r&d and extensive simulations will be necessary in order to determine the optimal module arrangement , the sensor technology ( e.g. the use of solid state detectors instead of pmt , see @xcite for a review ) , and other hardware related parameters .
it is worth mentioning an interesting outcome of this analysis : the method described here can be readily applied to existing and future icecube data in the regular muon data acquisition mode : events of @xmath206s duration are recorded at a rate of about @xmath118 khz , _
i.e. _ representing a lifetime fraction of about 5% , without the necessity of a dedicated data stream .
numbers provided in this paper can readily be rescaled to obtain the reach in the baseline icecube configuration and its deepcore provided a corresponding rescaling of the distance reach ( by a factor of about 4 ) . concerning a dedicated data stream
, we do not discuss here the possible hardware implications for the realistic implementations of the methodology , but it may well already be implementable within the current icecube data acquisition hardware .
the methodology is natural for operating @xcite and projected @xcite water detectors , given a comparatively large noise rate ( based on @xcite , we found in @xcite a discussion based on a similar approach ) .
we finally note that accounting for spin - flavor conversion and msw effect for neutrinos propagating outward the sn core @xcite , the @xmath207 flux is shifted to slightly higher average energy , improving the prospects of the coincidental hit method .
we acknowledge fruitful discussions with our colleagues in icecube , notably l. kpke .
this work is supported by the swiss national research foundation under the grant pp002114800 . | the current supernova detection technique used in icecube relies on the sudden deviation of the summed photomultiplier noise rate from its nominal value during the neutrino burst , making icecube a @xmath0 megaton effective detection volume - class supernova detector .
while galactic supernovae can be resolved with this technique , the supernova neutrino emission spectrum remains unconstrained and thus presents a limited potential for the topics related to supernova core collapse models .
the paper elaborates analytically on the capabilities of icecube to detect supernovae through the analysis of hits in the detector correlated in space and time .
these arise from supernova neutrinos interacting in the instrumented detector volume along single strings .
although the effective detection volume for such coincidental hits is much smaller ( @xmath1kton , about the scale of superk ) , a wealth of information is obtained due to the comparatively low coincidental noise rate .
we demonstrate that a neutrino flux from a core collapse supernova will produce a signature enabling the resolution of rough spectral features and , in the case of a strong signal , providing indication on its location .
we further discuss the enhanced potential of a rather modest detector extension , a denser array in the center of icecube , within our one dimensional analytic calculation framework .
such an extension would enable the exploration of the neutrino sky above a few gev and the detection of supernovae up to a few 100 s of kilo parsec .
however , a @xmath2mpc detection distance , necessary for routine supernova detection , demands a significant increase of the effective detection volume and can be obtained only with a more ambitious instrument , particularly the boosting of sensor parameters such as the quantum efficiency and light collection area . |
many exclusive @xmath0-decay amplitudes in the framework of scet contain @xmath0-meson distribution amplitudes @xcite .
the amplitude of the decay @xmath1 at large photon energies is given , up to power corrections , by a convolution of a hard part ( perturbatively calculable ) and the @xmath0-meson distribution amplitude @xcite .
amplitudes of some decays , e.g. , @xmath2 at large pion energies , contain both factorizable and non - factorizable contributions @xcite .
factorizable parts of decay amplitudes contain light - cone distribution amplitudes of the initial @xmath0-meson and final hadron(s ) .
they describe large - distance ( soft ) structure of these hadrons , and can not be calculated in perturbation theory .
the theory of hadronic distribution amplitudes in qcd is reviewed in @xcite .
quark antiquark distribution amplitudes of @xmath0-meson in hqet were introduced and investigated in @xcite .
they are defined as fourier transforms of matrix elements of some gauge - invariant bilocal operators between @xmath0-meson and vacuum .
renormalization of these operators was calculated in one - loop approximation .
however , an unusual term @xmath3 in the one - loop renormalization constants was erroneously omitted in @xcite .
the correct evolution equation for the leading - twist distribution amplitude was derived in @xcite at one loop .
the evolution kernel contains , in addition to terms obtained earlier @xcite , an unusual term @xmath4 .
the method of solution of the evolution equation is also discussed in @xcite in detail .
quark antiquark gluon distribution amplitudes of @xmath0-meson and their relations to quark antiquark ones ( based on equations of motion ) are discussed in @xcite .
sum rules for the quark
antiquark distribution amplitudes were obtained in @xcite . a simple model of these distribution amplitudes at a low normalization scale ( of order of hadronic scale ) was proposed .
radiative corrections to the perturbative term and the quark - condensate term were later calculated @xcite . in these lectures , we first briefly discuss what is hqet ( sect .
[ s : hqet ] ) .
a much more detailed presentation can be found in the textbooks @xcite . after a short discussion of @xmath5 ( sect [ s : fb ] ) , quark
antiquark distribution amplitudes are introduced ( sect .
[ s : qq ] ) .
quark antiquark
gluon distribution amplitudes and their relations to two - particle ones are discussed in sect .
[ s : qqg ] .
[ s : evol ] is the central ( and longest ) one . here
renormalization of light - cone bilocal quark operators in hqet is considered in one - loop approximation .
a detailed derivation of the evolution kernel is presented , based on the methods of @xcite .
finally , sum rules for the distribution amplitudes are briefly discussed in sect .
[ s : sr ] .
let s consider a heavy antiquark with mass @xmath6 and momentum @xmath7 in qcd . here
@xmath8 is a fixed 4-velocity ( @xmath9 ) , and the residual momentum @xmath10 , as well as momenta of all light quarks and gluons , are supposed to be small ( compared to @xmath6 ) .
the propagator of this heavy antiquark can be written as @xmath11 the leading term here is the hqet propagator .
this can be graphically presented as @xmath12{$mv+k$ } } \end{picture } } = \raisebox{-5.5mm}{\begin{picture}(22,8 ) \put(11,5){\makebox(0,0){\includegraphics{fr2.eps } } } \put(11,0){\makebox(0,0)[b]{$mv+k$ } } \end{picture } } + \mathcal{o}\left(\frac{1}{m}\right)\ , .
\label{graphprop}\ ] ] in a vertex sandwiched between such propagators , we may substitute @xmath13 projector can also be inserted near external legs , so that all qcd vertices can be replaced by the hqet ones , @xmath14{$\mu$ } } \put(8,9){\makebox(0,0)[t]{$a$ } } \end{picture } } = i g_0 t^a ( -v^\mu)\ , , \label{vertt}\ ] ] up to @xmath15 corrections .
these feynman rules can be obtained from the lagrangian @xmath16 the antiquark field is our main field here , it satisfies @xmath17 @xmath18 is the conjugate field for @xmath19 . here
the covariant derivatives are @xmath20 qcd tree diagrams are reproduced by hqet up to @xmath21 corrections ( they can also be reproduced , if we add the appropriate @xmath22 terms into the lagrangian )
. the heavy quark chromomagnetic moment is @xmath23 by dimensionality . at the leading order in @xmath22 ,
the heavy - quark spin does not interact with the gluon field .
therefore , it may be rotated at will , without changing the physics ( heavy - quark spin symmetry ) .
it may even be switched off ( superflavour symmetry ) .
we shall work with the spinless heavy antiquark , @xmath24 during most of these lectures , because this greatly simplifies reasoning and calculations . here
again @xmath25 is the main ( scalar ) field , and @xmath26 is its conjugate .
so , tree qcd diagrams , expanded in @xmath27 to some order , are reproduced by the corresponding hqet diagrams .
but what about loops ? here
things are not so simple @xcite .
let s consider , for example , the heavy light two - point diagram ( fig .
[ f : thresh ] ) with @xmath28 , where the residual momentum @xmath10 is small . by choosing @xmath8 along @xmath29
we can always ensure @xmath30 .
let s consider the integral @xmath31 ^ 2 ( -l^2)}\ , , \label{thresholdint}\ ] ] which has neither ultraviolet ( uv ) nor infrared ( ir ) divergences .
there are two regions of the loop momentum @xmath32 in this integral : * hard region @xmath33 . expanding the integrand in @xmath34 , @xmath32 , we have @xmath35 ^ 2 ( -l^2 ) } + 4 \frac{(m +
l \cdot v ) \omega}{\left[m^2-(mv+l)^2\right]^3 ( -l^2 ) } + \cdots \label{hardintegrand}\ ] ] * soft region @xmath36 . expanding the integrand in @xmath10 , @xmath37 , we have @xmath38 ^ 2 ( -l^2 ) } + 2 \frac{(k+l)^2}{\left[-2 m ( k+l ) \cdot v\right]^3 ( -l^2 ) } + \cdots \label{softintegrand}\ ] ] ( 40,11.5 ) ( 20,4)(0,0 ) ( 20,-4)(0,0)[b]@xmath39 ( 20,11.5)(0,0)[t]@xmath32 ( 7.5,-4)(0,0)[b]@xmath29 ( 32.5,-4)(0,0)[b]@xmath29 the contribution of the hard region is @xmath40 + \cdots \right]\,,\\ & \frac{\mu^{2\varepsilon } i_h}{\gamma(1+\varepsilon ) } = - \frac{1}{2\varepsilon } + \log\frac{m}{\mu } + \left ( \frac{1}{\varepsilon } - 2 \log\frac{m}{\mu } - 1 \right ) \frac{\omega}{m } + \cdots \end{split } \label{hardcontr}\ ] ] where the on - shell massive two - point integrals ( fig . [
f : os1 ] ) @xmath41 are @xmath42 this contribution is ir divergent .
( 32,11.5 ) ( 16,4)(0,0 ) ( 16,-4)(0,0)[b]@xmath43 ( 16,11.5)(0,0)[t]@xmath32 ( 3.5,-4)(0,0)[b]@xmath44 ( 28.5,-4)(0,0)[b]@xmath44 the contribution of the soft region is @xmath45 + \cdots \right]\,,\\ & \frac{\mu^{2\varepsilon } i_s}{\gamma(1+\varepsilon ) } = \frac{1}{2\varepsilon } - \log\frac{-2\omega}{\mu } - \left ( \frac{1}{\varepsilon } - 2 \log\frac{-2\omega}{\mu } - \frac{1}{2 } \right ) \frac{\omega}{m } + \cdots \end{split } \label{softcontr}\ ] ] where the hqet two - point integrals ( fig .
[ f : hqet1 ] , @xmath46 ) @xmath47 are @xmath48 this contribution is uv divergent .
( 32,11.5 ) ( 16,4)(0,0 ) ( 16,-4)(0,0)[b]@xmath49 ( 16,11.5)(0,0)[t]@xmath32 ( 3.5,-4)(0,0)[b]@xmath10 ( 28.5,-4)(0,0)[b]@xmath10 the complete result is finite : @xmath50 this expansion can be easily continued if desired .
what about higher loops ?
let s consider , for example , the two - loop two - point diagram with a small external residual momentum .
there are several regions of the loop momenta in this diagram ( fig .
[ f : regions ] ) . in each of them , some momenta are hard ( @xmath51 ) , some are soft ( @xmath52 ) .
as we have already seen , a heavy - quark line with a soft residual momentum becomes an hqet line .
soft massless lines are shown by dashed lines in the figure .
( 152,18 ) ( 16,9)(0,0 ) ( 56,9)(0,0 ) ( 96,9)(0,0 ) ( 136,9)(0,0 ) for example , in the second diagram in fig .
[ f : regions ] , the left loop is hard .
it contains a single scale @xmath6 .
all external momenta of this loop are soft , including those of the lines belonging to the other loop .
we can expand the integrand in these small momenta ; after taking the loop integral , we obtain a polynomial in these momenta . now from the point of view of the soft loop ( large distances ) this hard loop is just a local vertex .
this loop contains a single scale @xmath53 , and we can calculate it , obtaining a non - analytical function of the external residual energy @xmath53 . in a general multiloop diagram , hard lines must always form loops , so that momentum conservation can hold after neglecting all soft momenta .
there can be several disconnected hard parts ; each of them must contain at least one heavy line ( a subdiagram consisting of only light lines and having soft external momenta has no reason to be hard ) . from the point of view of the soft part ,
hard parts are just local vertices
. there may appear a soft subdiagram connected to the rest of the diagram only at such a vertex .
such a subdiagram is scaleless and hence vanish .
for example , we could consider one more region in fig .
[ f : regions ] : when the only soft line is the middle light line , all the rest are hard .
but this soft line forms a loop containing one local vertex ( the integrated hard loop ) , and hence it vanishes . in the usual hqet formalism ,
lagrangian contains local operators multiplied by matching coefficients .
qcd operators are also expanded in hqet operators with matching coefficients .
these matching coefficients are the only quantities in the theory which depend on the hard scale @xmath6 .
diagrammatically , they come from hard loops in qcd diagrams .
local operators produce vertices polynomial in their external momenta .
they appear in hqet diagrams , which contain only the soft scale @xmath53 .
these hqet diagrams are soft parts of qcd diagrams .
during these lectures , we shall mostly live in a world with a heavy antiquark having @xmath54 .
physics in the real world is the same , up to @xmath22 corrections .
we shall work in the @xmath8 rest frame .
the ground - state @xmath55-wave @xmath56 meson has the quantum numbers @xmath57 .
there are 2 @xmath58-wave excited mesons with @xmath59 and @xmath60 .
the heavy
light quark current @xmath61 has no definite parity ; the currents with parity @xmath62 are @xmath63 they have the quantum numbers of @xmath55-wave @xmath64 mesons and @xmath58-wave @xmath65 mesons .
the ground - state meson @xmath66 has a dirac wave function @xmath67 which satisfies @xmath68 and is normalized by @xmath69 .
the matrix element of @xmath70 from @xmath66 to vacuum is @xmath71 where the one - meson states are normalized by the non - relativistic condition @xmath72 the correlator of the heavy
light currents ( fig .
[ f : cor ] ) , @xmath73 has the structure @xmath74 therefore , the correlators of the currents with definite parity are @xmath75 the spectral density of , say , the correlator with @xmath76 is @xmath77 where @xmath78 is the residual energy of the ground - state meson , and the dots mean contribution of excited states .
( 56,24 ) ( 28,14)(0,0 ) ( 1,0)(0,0)[b]@xmath79 ( 55,0)(0,0)[b]@xmath80 now we shall return to the real world with @xmath59 heavy antiquark for a while , but still with @xmath81 .
the @xmath55-wave ground - state meson turns into a degenerate doublet with @xmath82 , @xmath83 .
the two @xmath58-wave mesons turn into two degenerate doublets : with @xmath54 , @xmath84 and with @xmath85 , @xmath86 .
all heavy light currents @xmath87 reduce , due to ( [ lowercompon ] ) , to 4 ones with @xmath88 the first 2 currents , with @xmath89 anticommuting with @xmath90 , have the quantum numbers of the ground - state @xmath91 , @xmath83 mesons ; the second 2 currents , with @xmath89 commuting with @xmath90 , have the quantum numbers of the @xmath58-wave @xmath92 , @xmath84 mesons . the correlators ( fig .
[ f : cor ] ) are @xmath93 or @xmath94 where @xmath76 for @xmath89 anticommuting with @xmath90 and @xmath95 for commuting . for @xmath96 and @xmath97 ,
the correlators are @xmath98 and @xmath99 .
their spectral densities are @xmath100 and @xmath101 , where @xmath102 and dots mean contribution of higher states .
therefore , @xmath103 the usual definition of the decay constants is @xmath104 where the relativistic normalization @xmath105 of single - meson states is used ( this normalization becomes meaningless in the limit @xmath106 , and therefore can not be used in hqet ) . comparing ( [ fb ] ) for @xmath0 or @xmath107 at rest with ( [ fb ] ) , we arrive at @xmath108 up to @xmath22 corrections .
let s consider the correlator ( [ cor5 ] ) with @xmath76 . equating the ground - state contributions to the spectral densities of the left - hand side and the right - hand one , we obtain @xmath109 therefore , @xmath110 where the matrix @xmath111 is defined up to a phase factor . we can re - write this result for the relativistic normalization of the meson state , in covariant notation : @xmath112 where @xmath113 ( of course , one can re - define the phases of @xmath114 and hence of @xmath115 ) .
after safely returning to the ideal world with a @xmath92 heavy antiquark , we want to invent an operator which probes more details of the structure of @xmath0-meson than the local current ( [ j ] ) . to this end
, we consider a bilocal gauge - invariant operator @xmath116 q(z)\ , , \label{ot0}\ ] ] where @xmath117 = p \exp \left [ - i \int_x^y a_\mu(z ) d z^\mu \right]\ , , \label{exy}\ ] ] and @xmath118 its matrix element from the ground - state meson to vacuum has 2 dirac structures : @xmath119 u\ , .
\label{phit}\ ] ] because @xmath120 . in
what follows , we shall often use light - front components of vectors .
let s introduce ( in the @xmath8 rest frame ) two light - like vectors @xmath121 light - front components of any vector @xmath122 are defined as @xmath123 we have @xmath124 in particular , @xmath125 we shall also use light - front components of @xmath126 : @xmath127 the definition ( [ phit ] ) can be re - written as @xmath128 u\ , .
\label{ot}\ ] ] if we introduce the operators @xmath129 then @xmath130 the @xmath0-meson distribution amplitudes are the fourier transforms of these functions : @xmath131 they are normalized by @xmath132 the function @xmath133 is the leading - twist distribution amplitude , and @xmath134 the subleading - twist one ( though there is no good definition of twist in hqet ) .
we can formally introduce the operators @xmath135 then @xmath136 the distribution amplitudes describe the distribution in the light - front component @xmath137 of the light - quark momentum in @xmath0-meson . the expansion of the operators ( [ opmt ] ) in @xmath138 reads @xmath139 or for matrix elements @xmath140 we can also reconstruct @xmath141 from @xmath142 : @xmath143 or for matrix elements @xmath144 ( you can easily check this by substituting ( [ on ] ) into ( [ intdn ] ) ; integration in @xmath145 yields @xmath146 ) .
the first moments @xmath147 can be found from the equations of motion .
the equation of motion for the heavy antiquark is @xmath148 and therefore we obtain @xmath149 the vector part has the structure @xmath150 the equation of motion of the light quark is @xmath151 and using @xmath152 we obtain @xmath153 finally , we arrive at @xmath154 let s now consider the second moments .
they involve two new ( nonperturbative ) hadronic parameters : @xmath155 where @xmath156\,,\quad \vec{h } = i \vec{d}\times\vec{d}\,,\quad \vec{\alpha } = \gamma^0\vec{\gamma}\,,\quad \vec{\sigma } = - \vec{\gamma } \gamma_5 \gamma^0\,.\ ] ] now we can calculate all matrix elements with 2 derivatives . from ( [ d0 ] ) and ( [ dvec ] ) , using the heavy - antiquark equation of motion ( [ heavyeom ] ) , we immediately find @xmath157 using the definition ( [ lambdaeh ] ) of @xmath158 , we immediately find @xmath159 the second spatial derivatives have the structure @xmath160 where the second coefficient follows from the definition ( [ lambdaeh ] ) of @xmath161 . we find @xmath162 using the light - quark equation of motion ( [ lighteom ] ) : @xmath163 u\ , .
\label{dij}\ ] ] finally , we arrive at @xmath164 what about @xmath0-meson distribution amplitudes in the real world with a @xmath65 heavy antiquark ?
it has 4 distribution amplitudes , as any pseudoscalar meson : @xmath165 \gamma_5 q(z)|b{>}_r = - i f_b m \tilde{\varphi}_p\,,\\ & { < } 0|{q^*_v}(0 ) [ 0,z ]
\gamma^\mu \gamma_5 q(z)|b{>}_r = f_b \left [ i \tilde{\varphi}_{a1 } p^\mu - m \tilde{\varphi}_{a2 } z^\mu \right]\,,\\ & { < } 0|{q^*_v}(0 ) [ 0,z ] \sigma^{\mu\nu } \gamma_5 q(z)|b{>}_r = i f_b \tilde{\varphi}_t ( p^\mu z^\nu - p^\nu z^\mu)\ , .
\end{split } \label{qcdps}\ ] ] similarly to ( [ trf ] ) , we have @xmath166 \gamma q(z)|m{>}_r = f \operatorname{tr}\gamma \left[\tilde{\varphi}_+ + \frac{\tilde{\varphi}_--\tilde{\varphi}_+}{2 t } \rlap / z\right ] \mathcal{m}\ , . \label{trfphi}\ ] ] therefore these 4 qcd distribution amplitudes can be expressed via 2 hqet ones : @xmath167 the qcd distribution amplitudes are usually considered as functions of the longitudinal momentum fraction @xmath168 .
it is usually assumed that @xmath169 at @xmath170 , and @xmath171 .
therefore , we shall assume that @xmath172 at @xmath173 , and @xmath174 .
the qcd distribution amplitudes are normalized as @xmath175 ( to derive the last formula , we used the @xmath138 expansion ( [ omn ] ) and the first moments ( [ mom1 ] ) ) .
the vector meson @xmath107 is described by 6 distribution amplitudes in qcd .
all 10 distribution amplitudes ( 4 for @xmath0 plus 6 for @xmath107 ) are expressed via 2 hqet distribution amplitudes @xmath176 .
this is a consequence of the heavy - quark spin symmetry .
we do nt present formulae for @xmath107 here ; they can be found in @xcite .
now we shall discuss relations of quark antiquark distribution amplitudes and quark antiquark gluon ones following from the equations of motion . in order to apply them ,
we need to differentiate with respect to the coordinates of the light quark and the heavy antiquark separately .
therefore , we go slightly off the light cone .
the generalization of ( [ phit ] ) to an arbitrary @xmath177 is @xmath178 u\ , , \label{phiz}\ ] ] where @xmath179 , and the fixed - point gauge @xmath180 is used to simplify notation . in order to apply the light - quark equation of motion ( [ lighteom ] )
, we apply @xmath181 to this definition .
the differentiation of the right - hand side is straightforward . in the left - hand side
we write @xmath182 the first two terms yield zero . in the fixed - point gauge @xmath183 where @xmath184 .
the matrix element of the operator containing @xmath185 contains 4 structures : @xmath186 i g_{\nu\mu}(uz ) z^\nu [ uz , z ] q(z)|m { > } = { } \\ & - f \left [ ( v_\mu \rlap / z - t \gamma_\mu ) ( \tilde{\psi}_a - \tilde{\psi}_v ) + i \sigma_{\mu\nu } z^\nu \tilde{\psi}_v - z_\mu \tilde{\psi}_x + \frac{z_\mu}{t } \rlap / z \tilde{\psi}_y \right ] u\ , , \end{split } \label{psi}\ ] ] and is parametrized by 4 quark antiquark
gluon distribution amplitudes ( note that this matrix elements vanishes when multiplied by @xmath187 ) . equating the coefficients of @xmath67 and @xmath188 , we obtain two consequences of the light - quark equation of motion : @xmath189 similarly , to use the heavy - antiquark equation of motion we apply @xmath190 the first two terms in the last operator yield zero .
now we move the heavy antiquark ; the center of the fixed - point gauge must be constant , and we have to use the gauge @xmath191 : @xmath192 we arrive at two consequences of the heavy - antiquark equation of motion : @xmath193 the functions @xmath194 are some new ( non - leading ) quark antiquark distribution amplitudes .
we are not interested in them here .
there are 2 equations involving only our familiar distribution amplitudes @xmath195 , namely , ( [ eom1 ] ) and a combination of ( [ eom2 ] ) and ( [ eom3 ] ) : @xmath196 the quark antiquark gluon distribution amplitudes in the momentum space are defined as @xmath197 performing fourier transform of the equations ( [ eom5 ] ) , we obtain @xmath198 where @xmath199\,,\\ & j(\omega ) = - 2 \frac{d}{d\omega } \int_0^\omega d\rho \int_{\omega-\rho}^\infty \frac{d\xi}{\xi } \left[\psi_a(\rho,\xi ) + \psi_x(\rho,\xi)\right ] - 4 \int_0^\omega d\rho \int_{\omega-\rho}^\infty \frac{d\xi}{\xi } \frac{\partial}{\partial\xi } \psi_v(\rho,\xi)\,.\end{aligned}\ ] ] if we insist on having @xmath200 , then the condition @xmath201 = 0 \label{cond0}\ ] ] must be satisfied .
it should follow from vanishing of @xmath202 at @xmath173 ( behaviour of the three - particle distribution amplitudes at the boundaries is discussed in @xcite ) .
if we new @xmath203 , we could solve the equations ( [ eomm ] ) for @xmath176 .
this is not very realistic , because we do nt know them .
the solution is a sum of two terms : @xmath204 where @xmath205 is the solution of ( [ eomm ] ) in the case if all quark antiquark gluon distribution amplitudes vanish ( it is called the wandzura wilczek part of the solution ) , and @xmath206 is induced by the gluonic terms .
the wandzura
wilczek part is @xmath207 ( fig .
[ f : ww ] ) ; it satisfies the normalization conditions ( [ norm ] ) .
note that @xmath208 is the maximum value of @xmath137 of the light quark if we assume that @xmath0-meson ( having the residual energy @xmath78 ) consists of the on - shell heavy antiquark ( always having zero residual energy ) and the on - shell light quark .
the gluon - induced part is given by some explicit integrals of @xmath203 ; we do nt present these long expressions here , they can be found in @xcite .
( 42,45 ) ( 21,24)(0,0 ) ( solid line ) and @xmath134 ( dashed line),title="fig : " ] ( 31,0)(0,0)[b]@xmath208 ( 41,0)(0,0)[b]@xmath53 the moments ( [ omn ] ) also consist of two contributions .
the wandzura
wilczek parts are @xmath209 the gluon - induced part does not contribute to the zeroth moments ( normalization ) and the first ones ; its contribution to the second moments ( [ mom2 ] ) is expressed via the normalizations of the quark
gluon distribution amplitudes : @xmath210 the contributions to the higher moments are expressed via the moments of @xmath203 ; the explicit formulae can be found in @xcite .
until now , we neglected renormalization of the considered operators . in this section
we shall discuss renormalization of the leading - twist @xmath0-meson distribution amplitude .
let s summarize what we know about the bare operators of interest .
there are 3 families of such operators , or 3 `` representations '' ( @xmath138 , @xmath53 , @xmath211 ) : @xmath212 q(z)\,,\\ & o_+(\omega ) = { q^*_v}(0 ) \gamma_+ \delta(i d_+ - \omega ) q(0)\,,\\ & o_+^{(n ) } = { q^*_v}(0 ) \gamma_+ ( i d_+)^n q(0)\ , .
\end{split } \label{bare}\ ] ] they can be converted into each other : @xmath213 as we shall see , not all of these relations survive renormalization .
we shall calculate matrix elements of these bare operators , therefore , we need the feynman rules for them . if we retain @xmath214 in all brackets @xmath215 in @xmath216 , we obtain the quark
antiquark vertex with @xmath217 shown in fig .
[ f : oqq ] .
when transformed to the @xmath53-representation ( [ barerel ] ) , this gives @xmath218 , as expected from the form ( [ bare ] ) of @xmath219 .
( 100,23 ) ( 16,10.5)(0,0 ) ( 16,23)(0,0)[t]@xmath211 ( 7,10)(0,0)@xmath29 ( 32,10)(0,0)[l]@xmath220 ( 66,10.5)(0,0 ) ( 66,23)(0,0)[t]@xmath53 ( 57,10)(0,0)@xmath29 ( 82,10)(0,0)[l]@xmath221 now let s retain a single @xmath222 in @xmath215 .
after some combinatorics , this gives @xmath223 ( i \partial_+)^{n - m}\ , .
\label{vqqg0}\ ] ] this gives the quark
gluon vertex of @xmath216 with @xmath224 shown in fig .
[ f : oqqg ] . transforming it to the @xmath53-representation ( [ barerel ] )
gives two @xmath225-functions ( fig .
[ f : oqqg ] ) .
of course , the integral in @xmath226 of this vertex vanishes , because @xmath227 does not interact with gluons .
( 100,25 ) ( 16,12.5)(0,0 ) ( 16,25)(0,0)[t]@xmath211 ( 16,0)(0,0)[b]@xmath228 ( 7,12)(0,0)@xmath29 ( 21,7)(0,0)@xmath10 ( 32,12)(0,0)[l]@xmath229 + ( 100,25 ) ( 16,12.5)(0,0 ) ( 16,25)(0,0)[t]@xmath53 ( 16,0)(0,0)[b]@xmath228 ( 7,12)(0,0)@xmath29 ( 21,7)(0,0)@xmath10 ( 32,12)(0,0)[l]@xmath230 the bare operators @xmath219 can be expressed via the renormalized operators @xmath231 ( @xmath232 is the @xmath233 renormalization scale ) , or vice versa : @xmath234 the renormalization `` matrices '' @xmath235 and @xmath236 are inverse to each other : @xmath237 the renormalized operators @xmath231 obey the renormalization - group equation @xmath238 where the anomalous dimension `` matrix '' is @xmath239 this equation can be derived in two ways : either we differentiate ( [ ren1 ] ) in @xmath240 and obtain 0 ( because the bare operator is @xmath232-independent ) , or we differentiate ( [ ren2 ] ) . at one loop @xmath241 where @xmath242 from ( [ anomdim ] ) we obtain @xmath243 the matrix element of @xmath219 between a state with the light quark with momentum @xmath29 and the heavy antiquark with momentum @xmath244 ( which are off - shell ) and vacuum is @xmath245\\ { } = { } & { < } 0|o_+(\omega;\mu)|q(p),{q^*_v}(p ' ) { > } + a_s \int z_+^{(1)}(\omega,\omega';\mu ) \delta(p_+-\omega ' ) \gamma_+ d\omega'\ , , \end{split } \label{m}\ ] ] where @xmath246 are the contributions of the one - loop diagrams shown in fig .
[ f : m ] , the matrix element of the renormalized operator is finite at @xmath247 , and @xmath248 contains only negative powers of @xmath249 ( in the term with @xmath250 , we may substitute @xmath251 instead of @xmath252 ) .
this allows us to find the renormalization `` matrix '' with one - loop accuracy .
( 106,42 ) ( 16,31.5)(0,0 ) between a quark
antiquark state and vacuum , title="fig : " ] ( 16,10.5)(0,0 ) between a quark
antiquark state and vacuum , title="fig : " ] ( 53,10.5)(0,0 ) between a quark
antiquark state and vacuum , title="fig : " ] ( 90,10.5)(0,0 ) between a quark
antiquark state and vacuum , title="fig : " ] let s calculate the first diagram ( fig . [ f:1 ] ) for @xmath253
it is @xmath254 \left[-k^2-i0\right]}\ , .
\label{m1a}\ ] ] the numerator can be simplified as @xmath255 ( 36,25 ) ( 18,13.5)(0,0 ) ( 18,25)(0,0)[t]@xmath53 ( 2,7)(0,0)@xmath29 ( 10,15)(0,0)@xmath10 ( 24,3)(0,0)@xmath256 this diagram can be written in the form @xmath257\ , , \label{m1b}\ ] ] where the term with @xmath258 comes from the second @xmath225-function in ( [ m1a ] ) .
this function is @xmath259 \left[-k^2-i0\right]}\,.\ ] ] now we use @xmath260 parametrization @xmath261 for both denominators , and also write the @xmath225-function as @xmath262 \frac{d\nu}{2\pi}\ , .
\label{delta}\ ] ] we obtain @xmath263\ , .
\end{split}\ ] ] shifting the integration momentum as @xmath264 to form a full square , we obtain @xmath265\\ & \quad{}\times \int \frac{d\nu}{2\pi } \exp\left[2 i \nu \left ( \frac{\alpha_1}{\alpha_1+\alpha_2 } \omega ' - \omega \right ) \right]\ , \int d^d k ' \exp\left[(\alpha_1+\alpha_2)(k^{\prime2}+i0)\right ] \end{split}\ ] ] the integral in @xmath266 is calculated using the wick rotation @xmath267 : @xmath268 the integral in @xmath269 gives a @xmath225-function .
we obtain an integral in two @xmath260-parameters : @xmath270\ , \delta\left(\frac{\alpha_1}{\alpha_1+\alpha_2}\omega'-\omega\right)\ , .
\end{split}\ ] ] it is always possible to calculate one integral in a `` radial '' variable @xmath271 in the space of @xmath260-parameters via @xmath89-function . in this case ,
the most convenient choice of such a variable is @xmath272 .
therefore , we insert @xmath273 under the integral sign , and substitute @xmath274 : @xmath275 the final result is @xmath276 functions @xmath277 which appear in the evolution kernel should be understood as distributions : they are always integrated with smooth test functions @xmath278 . the distribution @xmath279_+$ ] is defined by @xmath280_+ \varphi(\omega ' ) d\omega ' = \int f(\omega,\omega ' ) \left(\varphi(\omega')-\varphi(\omega)\right ) d\omega'\ , .
\label{plus}\ ] ] therefore , formally we can write @xmath281_+ + \delta(\omega-\omega ' ) \int f(\omega,\omega '' ) d\omega''\ , .
\label{plus2}\ ] ] the result ( [ m1b ] ) , ( [ f1 ] ) of the calculation of the diagram in fig . [ f:1 ] can be written via a @xmath282-distribution as @xmath283_+ + \delta(\omega-\omega ' ) \left ( \int f_1(\omega,\omega '' ) d\omega '' - \int f_1(\omega'',\omega ) d\omega '' \right ) \biggr]\ , .
\label{m1c}\ ] ] the coefficient of @xmath284 here can be calculated by substitution @xmath285 : @xmath286 substituting @xmath287 in the first integral , we have @xmath288 therefore , the final result for @xmath289 ( fig . [ f:1 ] ) is @xmath290\ , .
\label{m1}\ ] ] now we shall calculate the second diagram ( fig . [ f:2 ] ) for @xmath291 it is @xmath292 \left[-(p'-k)\cdot v - i0\right]}\ , .
\label{m2a}\ ] ] ( 36,25 ) ( 18,13.5)(0,0 ) ( 18,25)(0,0)[t]@xmath53 ( 36,10)(0,0)@xmath244 ( 33,16)(0,0)@xmath293 ( 15,3)(0,0)@xmath10 this diagram can be written in the form @xmath294\ , , \label{m2b}\ ] ] where the term with @xmath295 comes from the first @xmath225-function in ( [ m2a ] ) .
this function is @xmath296 \left[-(p'-k)\cdot v - i0\right]}\,.\ ] ] now we use @xmath260-parametrization ( [ alpha ] ) for both denominators ( it is convenient to multiply the linear denominator by 2 before this ) and ( [ delta ] ) for the @xmath225-function , and obtain @xmath297\,.\ ] ] shifting the integration momentum as @xmath298 to form a full square , we obtain @xmath299\\ & \quad{}\times \int \frac{d\nu}{2\pi } \exp\left[2
i \nu \left ( \frac{\alpha_2}{\alpha_1 } - \omega '' \right ) \right]\ , \int d^d k ' \exp\left[\alpha_1(k^{\prime2}+i0)\right]\ , . \end{split}\ ] ] taking the integrals in @xmath266 ( [ wick ] ) and @xmath269 , we obtain the integral in two @xmath260-parameters : @xmath300\ , \delta\left(\frac{\alpha_2}{\alpha_1}-\omega''\right)\,.\ ] ] now the best choice of the `` radial '' variable is @xmath301 , so we simply substitute @xmath302 : @xmath303 finally , @xmath304 now we rewrite @xmath305 ( [ m2b ] ) , ( [ f2 ] ) via a @xmath282-distribution : @xmath306_+ + \delta(\omega-\omega ' ) \left ( \int_0^\omega f_2(\omega-\omega '' ) d\omega '' - \int_0^\infty f_2(\omega '' ) d\omega '' \right ) \biggr]\ , . \label{m2c}\ ] ] the coefficient of @xmath284 is @xmath307 we introduced @xmath308 only to regularize possible infrared problems ; we only need the uv divergence of this diagram , which does not depend on @xmath308 .
therefore , we may assume @xmath309 .
the coefficient of @xmath284 is @xmath310 and the final result for @xmath305 ( fig .
[ f:2 ] ) is @xmath311\ , .
\label{m2}\ ] ] this one - loop diagram contains a @xmath3 uv divergence !
it is in the coefficient of @xmath284 given by the integral ( [ coefdelta ] ) .
the function @xmath312 has a @xmath313 uv divergence in the integration in transverse momentum ; the longitudinal integral ( [ coefdelta ] ) is again uv divergent ( [ coefdelta2 ] ) .
the operator at the vertex of fig .
[ f:2 ] is ( after integration in @xmath53 ) a light - like wilson line ; the heavy - antiquark propagator is a time - like wilson line .
the vertex correction to a time - like light - like cusp on a wilson line is known to have a @xmath3 uv divergence at one loop @xcite . the third diagram ( fig .
[ f:3 ] ) is @xmath314 \left[-k^2-i0\right ] \left[-(p'+p - k)\cdot v - i0\right]}\ , .
\label{m3a}\ ] ] the numerator can be simplified as @xmath315 we shall now demonstrate that this diagram is uv - finite and hence does not contribute to the renormalization constant .
( 36,25 ) ( 18,13.5)(0,0 ) ( 18,25)(0,0)[t]@xmath53 ( 2,7)(0,0)@xmath29 ( 10,15)(0,0)@xmath10 ( 36,10)(0,0)@xmath244 ( 36,16)(0,0)@xmath316 ( 18,3)(0,0)@xmath317 using @xmath260-parametrization ( [ alpha ] ) for the denominators and ( [ delta ] ) for the @xmath225-function , and obtain @xmath318\ , .
\end{split}\ ] ] shifting the integration momentum as @xmath319 we obtain @xmath320\ ; \int \frac{d^d k'}{(2\pi)^d } e^{(\alpha_1+\alpha_2)(k^{\prime2}+i0)}\\ { } = { } & 2 c_f \frac{g_0 ^ 2}{(4\pi)^{d/2 } } \omega \gamma_+ \int d\alpha_1\,d\alpha_2\,d\alpha_3\ , \delta\left(\frac{\alpha_1\omega'+\alpha_3}{\alpha_1+\alpha_2}-\omega\right ) e^{-a}\ , , \end{split}\ ] ] where @xmath321 inserting @xmath322 under the integral sign and making substitutions @xmath323 , @xmath324 , we have @xmath325 , and the integral in the `` radial variable '' @xmath271 is easily calculated : @xmath326 where @xmath327 y + ( \omega - \omega ' ) \frac{\omega}{\omega^{\prime2 } } p^2\,.\ ] ] the integral in @xmath328 is finite at @xmath247 ( and easy to calculate ) .
now we are ready to find the renormalization constant @xmath235 .
re - expressing @xmath329 in the matrix element ( [ m ] ) , we have @xmath330\ , . \end{split } \label{mren}\ ] ] the quark - field renormalization constants in qcd and hqet th the feynman gauge are @xmath331 finally , the renormalization constant `` matrix '' at one loop is @xmath332\ , .
\end{split } \label{z1res}\ ] ] the evolution kernel ( [ anomdim ] ) has the structure @xmath333\ , .
\delta(\omega-\omega')\ , .
\label{gammastr}\ ] ] the logarithm @xmath334 only appears linearly , to all orders of perturbation theory ; the coefficient of this logarithm is the cusp anomalous dimension @xcite : @xmath335 ( the two - loop term is also known @xcite ) .
the non - logarithmic part is written as @xmath336 , where the anomalous dimension of the local current @xmath70 ( [ j ] ) is @xmath337 ( the two- and three - loop terms are also known @xcite ) .
it determines the evolution of @xmath338 ; the difference @xmath339 appears in the evolution equation for @xmath340 .
the non-@xmath225 term is @xmath341 the evolution kernel has dimensionality of @xmath342 .
the evolution equation for the distribution amplitude is @xmath343 \varphi_+(\omega;\mu ) + \int \gamma(\omega,\omega';a_s ) \varphi_+(\omega';\mu ) d\omega ' = 0\ , .
\label{evol}\ ] ] how to solve it ?
powers @xmath344 are eigenfunctions of the integral operator in ( [ evol ] ) , by dimensionality : @xmath345 where @xmath346\ , , \label{tildegamma}\ ] ] from ( [ gamma ] ) ( here @xmath347 is the euler constant ) .
they are not eigenfunctions of the whole evolution operator , due to the logarithmic term . we can construct solutions with the power depending on @xmath232 : @xmath348 if the function @xmath349 obeys @xmath350 then @xmath351 will cancel in the evolution equation . dividing the definition ( [ xi1 ] ) by @xmath352 and integrating , we obtain @xmath353\ , , \label{xi}\ ] ]
where @xmath354 .
we shall use @xmath355 as an independent variable ( `` time '' ) in the evolution equation instead of @xmath232 . the function @xmath356 satisfies the evolution equation @xmath357 if @xmath358 at the leading order , @xmath359\,,\ ] ] and @xmath360 - \int_0^\xi \frac{\tilde{\gamma_0}(n+\xi)}{\gamma_0 } d\xi - \frac{\gamma_0}{\gamma_0 } \xi\,.\ ] ] here , from ( [ tildegamma ] ) , @xmath361 finally , at the leading order , @xmath362 - \xi } \frac{\gamma(1-n-\xi)\gamma(1+n)}{\gamma(1+n+\xi)\gamma(1-n ) } \exp\left[-\left(\frac{\gamma_0}{\gamma_0}+2\gamma_e\right)\xi\right]\ , .
\label{uxi}\ ] ] the distribution amplitude at @xmath363 can be expressed via its moments ( [ intdn2 ] ) : @xmath364 substituting the solutions ( [ solform ] ) instead of @xmath344 gives us the solution of the evolution equation ( [ evol ] ) with initial conditions . at the leading order ( [ uxi ] ) , @xmath365 } \exp\left[-\left(\frac{\gamma_0}{\gamma_0}+2\gamma_e\right)\xi\right ] \left(\frac{\omega}{\lambda_{\overline{\text{ms}}}}\right)^\xi\\ & { } \times \int_{-\infty}^{+\infty } { < } \omega^{-1-in}{>}_+^{(\mu_0)}\,\omega^{in}\ , \frac{\gamma(1-n-\xi)\gamma(1+n)}{\gamma(1+n+\xi)\gamma(1-n)}\ , \frac{dn}{2\pi}\ , .
\end{split } \label{sol}\ ] ] qualitatively , each `` bin '' in the distribution amplitude at @xmath363 near some finite @xmath366 produces , after evolution to a larger @xmath232 , a radiative tail , slowly decreasing ( as @xmath367 ) at @xmath368 , due to the evolution kernel @xmath369 ( which behaves as @xmath367 at large @xmath53 , by dimensionality ) .
therefore , the behaviour of the distribution amplitude at large @xmath53 is @xmath367 , up to logarithms .
the normalization integral ( [ norm ] ) of the distribution amplitude logarithmically diverges at large @xmath53 ; its moments ( [ omn ] ) with @xmath370 are power - divergent .
this means that the expression ( [ on ] ) for the local operators @xmath216 ( @xmath371 ) via @xmath219 is not valid for the renormalized operators .
renormalization of @xmath219 removes uv divergences in transverse momenta ; in order to renormalize @xmath216 , we should also remove longitudinal uv divergences ( at large @xmath53 ) . as an illustration @xcite ,
let s suppose that at a low @xmath363 the distribution amplitude @xmath133 is given by the simple model ( [ model ] ) .
namely , this initial condition is taken at the scale @xmath363 where @xmath372 where the quark model is supposed to work , so that the light quark has only momenta of order @xmath78 ( no radiative tail ) .
this function is shown by the solid line in fig .
( [ f : evol ] ) , where @xmath53 is measured in units of @xmath373 ( [ omega0 ] ) . using the leading - order solution ( [ sol ] ) of the evolution equation ( [ evol ] ) and supposing @xmath374 , we obtain the distribution amplitude at the scale where @xmath375 ( dashed line ) and @xmath376 ( dashed - dotted line ) .
we can see that the main part of the distribution amplitude ( at @xmath377 ) becomes lower and the radiative tail becomes more prominent when @xmath232 grows .
( 100,62 ) ( 50,31)(0,0)-meson distribution amplitude @xmath340,title="fig:",width=377 ]
in order to estimate the quark antiquark distribution amplitudes @xmath176 of @xmath0-meson from qcd sum rules , we consider the correlator of the local current @xmath378 ( [ jp ] ) ( having the quantum numbers of the ground - state meson ) and the bilocal operator @xmath379 ( [ opmt ] ) : @xmath380 we are most interested in its fourier transform @xmath381 analytically continuing it from @xmath382 to @xmath383 , we obtain @xmath384 where @xmath385 is the spectral density .
the correlator contains the contribution of the ground - state meson ( written explicitly in ( [ sr3 ] ) ) and of the continuum of excited states .
( 32,17 ) ( 16,9)(0,0 ) ( 1,0)(0,0)[b]@xmath386 ( 21,0)(0,0)[b]@xmath79 ( 31,17)(0,0)[t]@xmath387 for sufficiently small @xmath388 , we can calculate this correlator theoretically . the perturbative contribution
[ f : sr ] ) is given in the fixed - point gauge @xmath389 by the light - quark propagator from @xmath386 to @xmath387 : @xmath390 its fourier transform is @xmath391 inverting the laplace transform ( [ sr3 ] ) , we find the spectral densities ( the integration contour should be to the right of the singularity at @xmath392 ) : @xmath393 we obtain @xmath394 the perturbative spectral density @xmath395 describes the contribution of the on - shell quark antiquark intermediate state with energy @xmath249 into the correlator ( [ sr3 ] ) .
the on - shell heavy antiquark has zero energy ; the maximum value of @xmath137 of the on - shell light quark with energy @xmath249 is @xmath396 .
this explains the @xmath397-functions in ( [ rho0 ] ) .
the quark condensate contribution is also important here .
it contains the vacuum average @xmath398[0,z]q(z){>}\,.\ ] ] this non - collinear quark condensate can be expanded in terms of bilocal condensates @xcite , the leading term in this expansion is the bilocal quark condensate @xmath399q(x ) { > } = { < } \bar{q}q{>}\,f_s(x^2 ) \label{bilocal}\ ] ] with @xmath400 .
this leading term produces the same contribution into @xmath401 : @xmath402 the expansion of the bilocal quark condensate at small @xmath80 is @xmath403 where @xmath404 in general , it can be written as @xmath405 where @xmath406 has the meaning of the virtuality distribution function of vacuum quarks .
its moments are expressed via vacuum averages of local operators : @xmath407 in terms of this function , the quark - condensate contribution into the fourier - transformed correlators is @xmath408 in other words , the virtuality distribution function appears directly in the sum rules for the @xmath0-meson distribution amplitudes !
this is similar to the case of non - diagonal sum rules for the pion distribution amplitude @xcite .
the local operator expansion ( [ local ] ) gives @xmath409 of course , a smooth function can always be expanded in derivatives of @xmath225-function , but such an expansion does not tell us much about the shape of this function , unless we sum an infinite number of terms .
the bilocal quark condensate ( [ bilocal ] ) at large @xmath410 behaves as @xmath411 because the wilson line @xmath412 $ ] can be considered an hqet heavy - quark propagator , and this vacuum average is the correlator of two hqet heavy light currents , having @xmath0-meson as the lowest - energy intermediate state .
of course , the conditions ( [ momfs ] ) , ( [ asympt ] ) do nt determine the shape of the distribution function @xmath406 in a unique way .
they are satisfied by the bakulev
mikhailov ansatz @xcite @xmath413 where the parameters @xmath414 , @xmath415 are determined by ( [ momfs ] ) .
now we equate the theoretical result for the correlators to the result obtained from a phenomenological model of the spectral densities .
they have the contribution @xmath416 of the ground - state meson and that of the continuum of excited states . as usual
, this contribution is modeled by the perturbative spectral densities ( [ rho0 ] ) starting from a continuum threshold energy @xmath417 : @xmath418 with this model , the equality of the phenomenological expression for the correlator and the theoretical one , @xmath419 becomes @xmath420 where the perturbative spectral density is integrated over the `` duality interval '' of the ground - state meson ( from @xmath79 to the continuum threshold @xmath417 ) .
we obtain the sum rules for the @xmath0-meson distribution amplitudes @xmath421 where the functions @xmath422 describe the effect of subtracting continuum from the perturbative contribution when the spectral density is @xmath423 .
the perturbative contributions vanish at @xmath424 , because of the properties of the spectral densities ( [ rho0 ] ) . setting @xmath425 in the correlator ( [ sr1 ] ) , or integrating ( [ sr2 ] ) in @xmath226
, we obtain the well - known sum rule for @xmath426 : @xmath427 the natural energy scale in this sum rule is @xmath428 with @xmath429 , one finds the optimal value of the continuum threshold @xmath430 ; then a wide plato at @xmath431 $ ] exists , and yields @xmath432 .
we divide the sum rules ( [ sr ] ) by ( [ sr0 ] ) and take @xmath433 ( in the middle of the plato ) .
the results are shown in fig .
[ f : p ] , [ f : m ] , where @xmath53 is measured in units of @xmath10 .
they are automatically normalized ( [ norm ] ) .
of course , the leading - order sum rules can not tell us at what scale these distribution amplitudes are normalized ; this scale must be low , @xmath434 .
the perturbative contributions vanish at @xmath424 .
the quark - condensate contribution is the same for both @xmath133 and @xmath134 .
here we used the ansatz ( [ ansatz ] ) .
this contribution gives a sharp peak at low @xmath53 and @xmath435 . ] .
details of its shape are unknown , but it can not be wider because of the restriction on the first moment ( [ momfs ] ) .
this contribution falls off quickly at larger @xmath53 .
therefore , the distribution amplitudes at this low @xmath232 are only non - zero at @xmath377 , in accord with the expectations of the quark model .
( 100,62 ) ( 50,31)(0,0 ) ( solid line ) , perturbative contribution ( dashed line ) , and quark - condensate contribution ( dashed - dotted line),title="fig:",width=377 ] ( 100,62 ) ( 50,31)(0,0 ) ( solid line ) , perturbative contribution ( dashed line ) , and quark - condensate contribution ( dashed - dotted line),title="fig:",width=377 ] if we neglect the quark - condensate contribution ( though this is not a good idea ) and also the effect of continuum subtraction , then the perturbative contributions ( [ pert2 ] ) suggest the following simple model of distribution amplitudes : @xmath436 they are normalized ( [ norm ] ) ; from the first moments ( [ mom1 ] ) we obtain @xmath437 these functions ( fig .
[ f : model ] ) have something in common with the wandzura wilczek ones ( fig . [
f : ww ] ) , but in contrast to them they fall off smoothly at large @xmath53 .
( 42,36 ) ( 21,20)(0,0 ) ( solid line ) and @xmath134 ( dashed line),title="fig : " ] ( 9.5,0)(0,0)[b]@xmath373 radiative corrections to the perturbative spectral density and the dimension-3 quark - condensate contribution for @xmath438 were calculated in @xcite . with these corrections , one can check that the correlator ( [ sr2 ] ) satisfies the evolution equation ( [ rg ] ) .
the resulting sum rules should be used together with the sum rules for @xmath426 with radiative corrections @xcite .
the perturbative spectral density is @xmath439\\ x<1:&\displaystyle 2 c_f \frac{\alpha_s}{4\pi } \biggl [ 2 \bigl(\log(1-x)+x\bigr ) \log\frac{\omega}{\mu } + 2 \log^2 ( 1-x)\\ & \displaystyle { } + ( 2x-1 ) \log(1-x ) - x \biggr ] \end{array } \right . \end{split } \label{rho1}\ ] ] where @xmath440 .
now the spectral density does not vanish at @xmath441 , it is just suppressed by @xmath442 .
therefore , the perturbative contribution to the sum rule ( [ sr ] ) for @xmath133 does not vanish at @xmath424 .
it produces a radiative tail @xmath443 with the magnitude of order @xmath442 . the radiative correction to the dimension-3 quark - condensate contribution
can not be used together with a model of bilocal condensate , because radiative corrections to higher - dimensional contributions are not known .
it is vital to use some model of the bilocal quark condensate in the sum rules for the distribution amplitude , because the local operator expansion produces contributions @xmath444 , @xmath445 , , which do nt tell us much about the shape of the distribution amplitude .
therefore , the full result for this correction can not be used in the sum rule for @xmath133 .
fortunately , the authors of @xcite demonstrated that a part of this correction is universal , and exponentialized into the sudakov factor .
we may multiply the bilocal quark - condensate contribution by this sudakov factor .
i am grateful to m. neubert for collaboration on @xcite and to s.v .
mikhailov for useful discussions . | @xmath0-meson light - cone distribution amplitudes are discussed in these lectures in the framework of hqet . the evolution equation for the leading - twist distribution amplitude
is derived in one - loop approximation .
qcd sum rules for distribution amplitudes are discussed . |
massive stars have a profound affect on their natal environment .
they have strong ionizing radiation fields which create hii regions , and their powerful winds sweep up surrounding material creating wind - blown shells and cavities .
their supernovae ( sne ) chemically enrich the interstellar medium ( ism ) and help to sustain turbulence within it .
thus the presence of massive stars in stellar clusters has implications for future generations of star formation .
the dispersal and destruction of molecular material by the winds and ionzing radiation may inhibit further star formation within that region .
conversely , compression of material by winds and shocks may also trigger new star formation @xcite and new cluster formation @xcite .
many high - mass star - forming regions are observed to contain diffuse thermal x - ray emission , which requires high temperature plasma .
it has long been recognized that the fast winds of individual massive stars create high pressure and high temperature bubbles ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . in large clusters containing many early - type stars
the individual stellar winds may combine , collectively creating a so - called cluster wind ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . in many cases
the observed diffuse emission from massive - star forming regions ( msfrs ) is relatively soft .
for instance , in m17 and the rosette nebula the characteristic temperature @xmath7kev @xcite .
the x - ray emitting plasma in the extended orion nebula is similarly cool @xcite .
however , in other msfrs the characteristic temperature is considerably higher .
for instance , the diffuse thermal x - ray emission from ngc3603 @xcite , the core of the arches cluster @xcite , and the quintuplet cluster @xcite has @xmath8kev .
in some clusters the diffuse x - ray emission may be predominantly non - thermal : e.g. ngc6334 @xcite , rcw38 @xcite , the arches cluster @xcite and on2 @xcite .
this requires non - thermal particles , which also occur when high speed flows and strong shocks are present .
@xcite argue that the non - thermal emission detected from the ngc3576 ob association may arise from both a pulsar wind nebula and a cavity supernova .
a caveat to some of these works is that the diffuse x - ray emission from clusters which are at larger distances is more likely to suffer contributions from unresolved point sources .
a summary of x - ray studies of young stellar clusters is presented by @xcite .
early studies indicated that the detection of diffuse x - ray emission in stellar clusters required the presence of stars earlier than o6 @xcite , although an exception , the o7-powered hourglass nebula @xcite , was known . @xcite
have since discovered diffuse x - ray emission from ngc2024 ( the flame nebula ) , which contains only late o- to early b - type stars . on the other hand ,
it is curious that diffuse x - ray emission has yet to be detected from some very massive stellar clusters where very early o - type stars are present , such as trumpler16 @xcite .
other work has indicated that the temperature of the diffuse plasma may be correlated with how embedded it is , as measured by the column density @xcite
. clusters with high temperature plasma appear to have @xmath9 , while clusters with cooler plasma have less absorption .
@xcite suggest that while the stellar winds are bottled up the shocked gas remains maximally heated , but subsequent leakage and the resulting adiabatic cooling of the gas causes the gas temperature to drop .
unfortunately , past comparisons of x - ray observations with theory have had mixed success .
many works on msfrs simply compare the observed x - ray luminosity against the mechanical wind power of the stars , or the thermal energy of the plasma against an estimate of the time - integrated energy input of the winds ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the efficiency of the conversion of mechanical energy to radiation is then found to range from @xmath10 to 0.1 .
this , and the estimated mass of the x - ray emitting gas , indicates that in many cases the winds are not completely confined and that hot plasma must flow into the wider environment .
this conclusion is reinforced by the fact that the application of completely confined wind - blown - bubble models often leads to a significant overprediction of the x - ray luminosity ( e.g. * ? ? ?
* ; * ? ? ?
other works have compared the x - ray luminosity and the surface brightness profile of the diffuse emission to the predictions of cluster - wind models . in their analysis of the arches and quintuplet clusters
, @xcite found that the radial intensity profiles of the diffuse emission were more extended than theoretical predictions .
@xcite recently determined that the observed diffuse x - ray emission from the carina nebula was 60 times too faint compared to predictions from the @xcite model , and 10 times too luminous compared to the @xcite model .
this led @xcite to develop a third model whereby density variations in the ism surrounding the cluster causes gaps in the swept - up shell , through which some of the high pressure gas in the bubble interior can leak .
their new model predicts a lower pressure within the bubble than the @xcite model , as the wind material is not completely confined , and also a lower x - ray luminosity .
consequently , it is more consistent with observations .
however , since the covering fraction of the shell is a free parameter this model suffers from a lack of predictive power . in young msfrs
in which there has not yet been time for any massive star to explode as a supernova , the diffuse x - ray emission must result from the action of stellar winds . however , in older clusters where some of the massive stars have exploded one still might not detect any signature of a sn explosion because the effect of a snr on the thermal properties of the hot cluster gas is likely to be relatively short - lived .
this time scale is generally believed to be @xmath11yr ( e.g. * ? ? ?
for this reason , most studies of stellar clusters prefer a wind based explanation for the diffuse x - ray emission , though @xcite favour a recent sn explosion in their study of the eastern tip of the carina nebula .
a distinction exists between individual stellar clusters , and larger scale regions of star formation which create superbubbles where multiple cavity supernovae are believed to be responsible for the diffuse emission ( such as those of 30doradus , e.g. * ? ? ?
* ; * ? ? ?
given the challenges of interpreting such complex environments as msfrs , and the highly idealized models of most theoretical and modelling work , in this paper the hydrodynamical models of stellar wind and supernova feedback in an inhomogeneous environment outlined in @xcite ( henceforth referred to as paper i ) are used as a basis to simulate the resulting x - ray emission from such regions .
of great interest are the x - ray luminosity and spectrum , and their temporal variation as the stars in the simulation cycle through various evolutionary stages , including main sequence , red supergiant , wolf - rayet and supernova . in section [ sec : xray_model_details ] the details of the model and the method of calculating the x - ray emission and absorption are discussed .
the results are presented in section [ sec : xray_results ] .
comparisons to numerical models and observations are made in sections [ sec : comp_theory ] and [ sec : comp_obs ] respectively .
section [ sec : xray_conclusions ] summarises and concludes this work .
the x - ray calculations in this paper are based on the 3d hydrodynamical model described in paper i. the simulations were performed using the hydrodynamical code arwen , which uses a piecewise parabolic interpolation and characteristic tracing to obtain the time - averaged fluid variables at each zone interface .
an iterative riemann solver is used to determine the time - averaged fluxes and solve the equations of hydrodynamics ( see paper i for more details ) .
the simulations consist of three massive o stars which represent the main sources of mechanical feedback in a massive star forming region contained within an inhomogeneous gmc clump of radius 4pc and mass 3240m@xmath0 .
the medium surrounding this clump is homogeneous , with a density of 3.33@xmath12 ( @xmath13 ) and a temperature of 8000k .
the simulations were performed on a 512@xmath14 grid with free outflow boundary conditions .
the total simulation volume covers a cubic region of @xmath1516pc centered on the gmc clump .
the cluster wind is injected as thermal energy within a radius of 6 cells ( 0.375pc ) .
the evolution of the three stars is treated simplistically as three distinct phases - the main sequence ( ms ) , red supergiant ( rsg ) and wolf - rayet ( wr ) phases .
the details of the stellar cluster are summarized in table [ evolution ] . at the end of the wolf - rayet phase
the stars explode imparting 10m@xmath0 of material and 10@xmath16ergs of thermal energy into the environment .
the lifespans of the stars are designed in such a way so that there are three distinct supernova explosions over the course of the simulation .
the simulation utilizes a temperature dependent average particle mass , and molecular , atomic and ionized phases are tracked separately .
the net heating / cooling rate per unit volume is parameterized as @xmath17 , where @xmath18 .
@xmath19 is the heating coefficient and is set at a constant value of @xmath20 .
@xmath21 is the cooling coefficient which is assumed to depend only on temperature .
cooling at low temperatures ( t@xmath22k ) is then adjusted to provide three thermally stable phases which correspond to the molecular ( t@xmath23 10k ) , atomic ( t@xmath23 150k ) and ionized ( t@xmath23 8500k ) phases .
this cooling curve and the phase diagram are shown in @xcite .
photoevaporation is not treated in these simulations .
however , as the photoevaporation time for a clump of these characteristics is comparable to the lifetime from ablation this should not significantly affect the results .
the simulations in paper i showed that the inhomogeneous structure of the natal gmc cloud surrounding the cluster had an important effect on the initial expansion of the cluster wind , which cut channels through the low density material to escape the clump .
the regions of high density within the initial clump proved to be surprisingly resistant to ablation from the cluster wind , and at later times the shockwaves of the snrs .
.wind properties of the three stars in the cluster as they evolve .
[ cols="^,^,^,^,^,^ " , ] 1e0102.2 - 7219 ( hereafter 1e0102 ) is a snr in the small magellanic cloud ( smc ) with an inferred kinematic age of @xmath232100yrs @xcite at a distance of @xmath2360kpc @xcite .
the progenitor is thought to have been a wolf - rayet star with a zero - age ms ( zams ) mass of @xmath2332m@xmath0 that underwent significant mass loss prior to exploding as a type ib / c or iil / b supernova .
@xcite used _ chandra _ to image the snr , finding it to be almost `` textbook '' , with a hotter outer ring surrounding a cooler , denser inner ring which is likely the reverse - shocked stellar ejecta .
the diameter of the snr was estimated to be 40 by @xcite ( approximately 12pc at 60kpc distance ) .
more recently , @xcite estimated the radius of the blast wave to be 6.4pc , in good agreement with their earlier work . @xcite
also estimated an expansion age of @xmath231000yrs , though @xcite disagree , deriving a free expansion age of 2100yrs .
however , recently @xcite have proposed that the source is a be / ns binary .
there is little interstellar extinction along the line of sight to 1e0102 which allows a comprehensive , multi - wavelength analysis from the x - ray to the radio domain .
@xcite estimated the expansion velocity of the blast wave to be 6000kms@xmath3 based on a radius of 6pc and an age of 1000yrs , although @xcite find from doppler shifts that the majority of the bulk matter is moving at a lower 1000kms@xmath3 .
@xcite also estimate the temperature in the postshock region to be 0.41.0kev .
@xcite estimated the upper limit on x - ray emission of the central source to be @xmath24 , whilst @xcite estimate the 0.510kev luminosity of the be / ns binary to be @xmath25 using suzaku data .
this is only a factor of 2 lower than our intrinsic 0.510kev x - ray emission from the simulated cluster , which is l@xmath26 and l@xmath27 for 1000 and 2000yrs after the first sn explosion respectively .
as the simulation assumes collisional ionization equilibrium ( cie ) , which is unlikely to be the case in 1e0102 , it is not surprising that higher luminosities are obtained from the simulation .
the clumpy environment in the simulated cluster may also be partially responsible , if the environment of 1e0102 has lower density and/or is more homogeneous .
the expansion velocity of the simulated blastwave is between @xmath2360008000kms@xmath3 ( see table [ snecomparison ] ) , which although high is similar to the estimate of the blast wave by @xcite .
also known as g292.0 + 1.8 , this is a core - collapse sn with an estimated age of 2700 - 3200years @xcite , and a distance of 6kpc @xcite .
it is one of only a handfull of o - rich snrs known today @xcite .
the x - ray emission from such o - rich snrs is thought to arise from faster , non - radiative shocks in lower density ejecta and interstellar gas .
the central source is thought to be a pulsar wind nebula .
the snr has a radius of approximately 15pc assuming a distance of 6.2kpc @xcite .
@xcite derived an average temperature for the snr using two components - a high temperature plasma associated with the supernova blast wave and a low temperature plasma from the reverse shock .
these two components were estimated to be 1.05 @xmath15 0.34 and 0.37 @xmath15 0.18kev respectively .
the progenitor star is estimated to have had a mass of 30 - 40m@xmath0 , though @xcite estimated a lower mass of 20 - 25m@xmath0 .
a more recent estimate of the temperature by @xcite using _ chandra _ data found a highly non - uniform distribution of hot , x - ray emitting gas in the remnant ranging from kt@xmath235kev in the nw regions to around [email protected] in the se .
these results are a promising match with our results 20003300years after the explosion of the 35m@xmath0 star , when the average temperature of the x - ray emitting plasma is @xmath280.540.68kev . the sn explosion producing msh11 - 54 is thought to have been asymmetric , which would explain the spatial variation of temperatures and the greater expansion of the remnant towards the nw .
@xcite found the unabsorbed 0.310kev x - ray luminosity of the central pulsar to be l@xmath29 , but made no estimate for the entire remnant .
our simulated cluster at around this time has a total 0.110kev intrinsic luminosity of l@xmath30 .
n132d is one of the brightest snrs in the large magellanic cloud ( lmc ) and has an estimated age of 3150years @xcite and an inferred progenitor mass of 30 - 35m@xmath0 @xcite . with a diameter of 80
, the distance to the snr of approximately 55kpc @xcite implies a real diameter of @xmath2321pc . this is a similar estimate to the extent of the x - ray shell , which has an estimated radius of 12pc @xcite .
the expansion velocity of the snr has been estimated by several authors ( e.g. * ? ? ?
* ; * ? ? ?
* ) , with values ranging from 22503700kms@xmath3 . in comparison ,
the radius of our simulated snr at around 3300yrs after the explosion is @xmath31pc , with an inferred expansion velocity of @xmath234400kms@xmath3 .
the x - ray luminosity in the 0.24kev energy band was estimated by @xcite to be 4.57.5@xmath32 , based on thermal plasma temperatures of 10@xmath33k and a hydrogen column density of 10@xmath34 @xcite .
the estimated luminosity is actually higher than the simulated results ( l@xmath35 at 3300yrs ) , despite the fact that the simulations assume cie .
however , given the inherrent differences in the simulated and actual environments these results can be considered a reasonable match .
the plasma temperature is very similar to that found in the other snrs mentioned , at approximately 0.6 - 0.7kev , compared with an average of 0.54kev from our simulated remnant ( see table [ xraygasprops ] ) .
there is patchy x - ray absorption around the remnant thought to be caused by gas just outside the molecular cloud towards the nothern tip of n132d @xcite .
puppis a is a nearby galactic snr and has age estimates ranging from 3700yr @xcite to 4450yr @xcite , making it most comparable to the simualation 3900yrs after the explosion ( see table [ snecomparison ] ) .
a distance of 2.2kpc has been estimated based on hi and co studies @xcite , although a closer distance of 1.3kpc has also been proposed by @xcite based on oh line detections .
this remnant is embedded in a complex region composed of large atomic and molecular clouds and an interstellar density gradient .
the remnant is about 50 in diameter ( approximately 32pc at a distance of 2.2kpc ) .
a progenitor mass of 25@xmath36 was inferred by @xcite .
optical knots detected from puppis a are evident only in the northeast , implying the ejection of the matter during the explosion was asymmetric @xcite .
oxygen - rich filaments are detected to have radial velocities higher than @xmath231500kms@xmath3 .
these filaments are interpreted as sn ejecta which have remained mostly uncontaminated by the ism @xcite .
recently , @xcite studied puppis a using _
chandra _ and xmm - newton .
they estimated the x - ray luminosity between 0.3 and 8.0kev to be l@xmath37 assuming a distance of 2.2kpc .
the x - ray emission from puppis a appears to be dominated by the swept - up ism due to very low metal abundances @xcite .
the total intrinsic x - ray luminosity of our simulated remant 3860yrs after the explosion is l@xmath38 , which is a reasonable match to puppis a. the average temperature in the remnant is 0.6kev , very similar to the average temperature of 0.54kev seen in the simulated remnant at 3860years after the explosion ( see table [ xraygasprops ] ) .
this paper investigates the x - ray emission from a massive young stellar cluster embedded in an inhomogeneous gmc clump treating only the mechanical effects from winds and supernovae . the hydrodynamical input model was previously simulated in paper i , and this work explores the emission arising from that model .
initially the dense parts of the clump decrease the observed x - ray emission due to attenuation , but once the cluster wind has destroyed and ablated a large portion of this material the attenuation from the ism material is dominant . at very early times , when the wind material is still confined by the inhomogeneous gmc material the x - ray luminosity is reasonably bright , at l@xmath39 .
however , as the cluster wind errodes and destroys the surrounding clump it is no longer completely confined and therefore hot gas is able to leak through the gaps in the shell .
this causes a reduction in the x - ray luminosity as the pressure within the bubble decreases .
once the low density gas from the clump has been ablated away the covering factor of the cluster remains more or less constant , leading to an approximately constant intrinsic x - ray luminosity of 1.7@xmath4 and an attenuated x - ray luminosity of 7@xmath40 .
the most massive star becomes a rsg at t=4.0myrs , resulting in a large drop in the x - ray luminosity in all three of the x - ray broadbands studied .
the most dramatic decrease is seen in the bb3 ( 2.510.0kev ) emission , where the attenuated x - ray luminosity drops four orders of magnitude , from l@xmath41 to l@xmath42 by the end of the rsg phase .
the drop in x - ray luminosity in the other two broadbands over this period is around a factor of 50 and 100 for bb1 ( 0.10.5kev ) and bb2 ( 0/52.5kev ) respectively .
although a lot of material is deposited in the rsg - enhanced cluster wind , the amount of material at x - ray emitting temperatures is very low , contributing to the lack of x - ray emission observed at this time .
100,000years later the most massive star further evolves to become a wr star , causing a dramatic increase in x - ray emission in all three broadband regions studied .
the amount of material at a temperature greater than 10@xmath43 increases by an order of magnitude over that seen in the rsg stage , and a total of 78@xmath44 of the computational volume contains x - ray emitting material .
the high momentum cluster wind sweeps up the slower moving material deposited in the previous phase and heats it to high temperatures , with the average temperature of hot gas ( @xmath45k ) at this time being around t@xmath46k .
the total attenuated x - ray emission increases to l@xmath47 , which is about 30 times greater than that observed when all three stars were on the ms . at t=4.4myrs the most massive star in our
simulation explodes as a sn , ejecting 10@xmath36 of material and 10@xmath48 of energy into the simulation . due to the way in which the explosion is initialised
, the emission from the snr only becomes independent of the initial conditions of the explosion once the interaction of the blastwave with the surrounding material becomes dominant . in this work
this occurs approximately 900years after the explosion .
the ejecta begins to leave the grid 4600years after the explosion , and therefore the emission from the snr can be compared with observations of snrs only between the ages of @xmath49 years .
the supernova of the 35m@xmath0 star was compared with four young core - collapse sne with ages ranging from @xmath2310004450yrs . although collisional ionization equilibrium was assumed in our simulation , which is unlikely to be true for such young remnants , we find that the x - ray luminosity and electron temperatures are reasonable matches to observational results reported in the literature . unfortunately , as the ejecta begins to leave the grid 4,600yrs after the initial explosion no comparisons can be made with older remnants .
the simulated emission from the cluster during the wind - dominated phases is substantially lower than predicted by 1d
spherically symmetric wwb calculations , but is higher than predicted by cluster wind models .
the fact that our simulations fall between these two models ties in very nicely with the theory of leaky bubble models , where the wbb is only partially confined .
the simulated results match reasonably well to actual observations of several massive young stellar clusters .
this is likely to be due the assumptions made in these calculations being overly simplified compared to the simulated model .
firstly , the assumption that the hot wind material is confined within a bubble is very much not the case , and a reduction in the pressure in the simulated cluster leads to a reduction in the x - ray luminosity .
the surrounding density is not homogeneous as described in these models , which will lead to local areas of confinement and leakage . clearly
the highly complex environment of young massive star forming regions requires similar complexity in simulations in order to better understand their properties .
hr acknowledges a henry ellison scholarship from the university of leeds and jmp acknowledges funding from the royal society for a university research fellowship and from the stfc .
we would also like to thank the referee for their timely and useful comments .
rcw38 is a very young ( @xmath50myr ) , highly embedded ( @xmath51 ) , and close ( @xmath52kpc ) stellar cluster surrounding a central pair of o5.5 stars ( irs2 ) ( e.g. * ? ? ?
the dominant irs2 stars have cleared a region completely free of dust out to a radius of 0.1pc .
the extinction is patchy , and the hii region appears to be breaking out of the surrounding molecular gas in some directions . extended warm dust is found throughout a @xmath53pc region and coincides with extended ( @xmath54pc ) x - ray plasma which is predominantly non - thermal @xcite . the power - law index of the emission steepens toward the cluster core .
contamination of the diffuse emission by unresolved point - sources is not significant at distances of more than 0.15pc ( @xmath55arcseconds ) from the cluster center , though may be responsible for the more thermal nature of the diffuse emission measured in the core @xcite .
the cause of the non - thermal emission remains unclear .
the diffuse emission is strongest in the central core near irs2 , and is confined on the southeast along a ridge .
recent _ spitzer _
observations reveal that winds from irs2 have caused outflows towards the northeast , northwest and southwest of the central cluster ( see e.g. fig . 4 in * ?
* ) .
m17 is a very young blister h ii region located on the northeast edge of one of the largest giant molecular clouds in the galaxy , with an extent of @xmath57 ( @xmath58pc , * ? ? ?
the geometry of m17 is thought to resemble the orion nebula hii region except that it is seen edge - on rather than face - on @xcite .
m17 is photoionized by the massive stellar cluster ngc6618 , which has 14 identified o stars @xcite , and is estimated to be @xmath59myr old @xcite .
the distance has recently been determined using trigonometric parallax to be 2.0kpc @xcite .
several obscured o4-o5 stars form a central 1 arcminute ring in ngc6618 , and are principally responsible for ionizing the nebula .
extinction is patchy ( @xmath60 with an average of 8 to the ob stars , though some parts of the cluster have @xmath61 , * ? ? ?
the earliest o stars are an o4+o4 visual binary known as kleinmann s anonymous star @xcite , which may in fact be a pair of colliding wind binaries @xcite .
evidence for an older ( @xmath62myr ) stellar population to the north is presented by @xcite .
the diffuse x - ray emisison from m17 has previously been analyzed by @xcite , @xcite , @xcite , and most recently by @xcite .
it has relatively high surface brightness and blows out to the east of the cluster , extending nearly 10pc from the cluster .
the plasma appears to be channelled by the famous northern and southern ionization bars @xcite , and maintains roughly constant temperature as it flows @xcite .
although obscuration changes across the field , a global fit to the diffuse emission with a 3-temperature nei model yields @xmath63 , @xmath64 and @xmath65kev , with the highest temperature component providing 56% of the intrinsic luminosity @xcite .
the absorption to each of these emission components increases with the temperature of the component , so that the @xmath66 component suffers 6 times as much obscuration as @xmath67 .
there are indications that the shocked gas is not in complete ionization equilibrium , which is suggestive of it recently being shocked .
several gaussian lines are also needed - the cause is speculated to be charge exchange processes .
the total x - ray luminosity is @xmath68 .
@xcite previously determined that the x - ray plasma had a mass of @xmath69 , which when rescaled to a distance of 2.1kpc becomes @xmath70 ( for an assumed distance @xmath71 , @xmath72 , @xmath73 , @xmath74 , and @xmath75 ) .
@xcite determine that @xmath76 .
the analysis by @xcite , which does not quite include the most easterly extent of the plasma , is generally consistent with the earlier work of @xcite , except for the determination of a significantly lower plasma temperature of @xmath77kev .
westerlund2 ( hereafter w2 ) is a compact young open cluster embedded in and responsible for the luminous hii region rcw49 .
w2 contains at least a dozen ob stars .
two wr stars , wr20a and ( especially ) wr20b , lie outside the cluster core ( see references in * ? ? ?
w2 is also located in the direction of one of the galaxy s strongest sources of @xmath78-rays @xcite .
the distance to w2 has been very difficult to pin down , with estimates ranging from 2 to more than 8kpc in the literature , but a new study puts it at @xmath79kpc , and determines an age of no more than 2myr @xcite .
diffuse x - ray emission from w2 was identified in a _
observation @xcite .
the emission is brightest at the core of w2 , and extends preferentially towards the west .
the emission can be fitted with a 3-temperature thermal plasma model with @xmath80 , @xmath81 and @xmath82kev , with the highest temperature component providing 30% of the intrinsic luminosity @xcite . assuming a distance of 2.3kpc , the absorption corrected 0.5 - 8kev luminosity is @xmath83 . at @xmath84kpc ,
this increases to @xmath85 .
the absorbing column to @xmath67 is less than the identical columns to @xmath86 and @xmath66 .
the hardest thermal component is not well constrained , and replacing it with a power - law component ( @xmath87 ) also gives an acceptable fit .
the diffuse flux will be slightly underestimated due to the use of a 5arcmin radius extraction region and a nearby on - chip background region .
more recently diffuse emission has also been analyzed from a _ suzaku _ observation @xcite .
pointing was used to determine the point source contamination to the _
suzaku_-derived diffuse emission , and the central region ( @xmath88arcmin ) was masked out .
the diffuse emission is found to extend to an 8 arcminute radius .
the spectral analysis is broadly consistent with the earlier results of @xcite .
the orion nebula cluster ( onc ) , also known as the trapezium cluster , contains the nearest rich and concentrated sample of pre - ms stars .
the ob members of the onc photoionize the orion nebula ( m42 ) , a blister hii region at the near edge of orion a , the nearest giant molecular cloud ( @xmath89pc ) .
@xcite recently detected diffuse , soft ( @xmath90kev ) x - ray emission in the extended orion nebula ( eon ) .
the characteristic temperature of the plasma is @xmath91kev .
the intrinsic x - ray luminosity in the @xmath92kev energy band is @xmath93 .
two regions of diffuse emission , a northern and a southern , are identified with respective emission measures of @xmath94 and @xmath95 .
the attenuating column @xmath96 is very low , being @xmath97 for the northern , and @xmath98 for the southern . the total mass of the x - ray emitting gas is estimated to be @xmath99 , which is roughly @xmath100yrs of mass - loss of the dominant o5.5 star @xmath101oric .
the radiative cooling time is estimated to be @xmath102myr .
the density of the x - ray plasma is determined to be @xmath103 .
the x - ray and ionized gas are in approximate pressure equilibrium ( @xmath104 ) , and the hot gas is likely channeled by the cooler denser structures rather than disrupting them by expansion .
leakage of the hot plasma via an x - ray champagne flow into the nearby eridanus superbubble is suggested .
the rosette nebula is a blister hii region at the tip of the giant rosette molecular cloud .
it has a distinct ringlike appearance in both radio and optical images , and is photoionized by the open cluster ngc2244 whose stellar winds have cleared a hole in the nebula s centre @xcite .
ngc2244 contains 7 o - type stars , all of which have ms luminosity classes , with the earliest spectral type being o4v((f ) ) . a recent analysis of 6 of these stars by @xcite determined an upper age limit of 2myr for the most massive stars , in excellent agreement with earlier determinations ( e.g. *
* ; * ? ? ?
photometric distance estimates range between 1.4 and 1.7kpc , and 1.55kpc is adopted in this work .
@xcite find an absence of mass segregation and conclude that the cluster is not dynamically evolved .
the two dominant o stars ( hd46223 , o4v((f ) ) ; hd46150 , o5v((f))z ) are widely separated ( by at least 3pc ) . in contrast , the o stars in the trapezium group and m17 are concentrated within the inner 0.5pc .
@xcite find that soft diffuse x - ray plasma surrounds the ob association and fills the nebula cavity completely .
it likely originates from the o - star winds which are thermalized by wind - wind interactions or by shocking against surrounding molecular material .
the x - ray emission is brightest in the central 3pc radius , corresponding roughly to the central cavity .
the diffuse emission can be fit by a two - temperature thermal plasma model , with components @xmath105 and @xmath106kev and a single absorbing column @xmath107 .
the hotter component is dominant .
the intrinsic @xmath108kev luminosity ( for @xmath109kpc ) is @xmath110 .
there is no significant emission above 2kev .
correcting @xcite s values for a slightly greater assumed distance , the diffuse plasma number density and mass are estimated as @xmath111 and @xmath112 .
the quintuplet cluster is named after its five brightest stars @xcite .
it is located near the galactic centre , is unusually dense , and is host to at least 10 massive , windy , wr stars and more than a dozen luminous ob supergiants @xcite .
it is somewhat less massive and dense than the arches cluster , however .
its age is estimated to be about @xmath113myr @xcite .
@xcite determine that the quintuplet cluster shows thermal diffuse emission with a peak temperature @xmath114kev at the cluster centre , and with @xmath115 .
the diffuse emission is much fainter than that in the arches and has a very low surface brightness .
it also has essentially the same spectral shape as the integrated spectrum from the detected sources .
considering the distance to the cluster , contamination by unresolved point - sources may be an issue .
@xcite analyze a deeper _ chandra _ exposure .
they report the same concerns as @xcite and in addition note that the extent of the diffuse emission from the quintuplet cluster is uncertain . with an extraction radius of 1arcmin ,
@xcite find that a single - temperature thermal plasma model yields @xmath116kev and @xmath117 , giving an absorption - corrected @xmath118kev luminosity of @xmath119 .
the radial diffuse x - ray intensity profile falls off more rapidly than sph simulations @xcite predict .
westerlund1 ( hereafter w1 ) is the most massive stellar cluster known in the galaxy @xcite .
it contains a rich population of massive stars which include more than 20 wr stars @xcite , more than 80 ob stars , and short - lived transitional objects including luminous blue variables ( lbvs ) , red supergiants ( rsgs ) and half the currently known population of yellow hypergiants ( yhgs ) in the galaxy .
estimates for its age range from @xmath120myr @xcite to @xmath121myr @xcite .
its distance remains somewhat uncertain , but estimates appear to be converging on the range @xmath122kpc ( see * ? ? ?
* and references therein ) .
it shows evidence of mass segregation @xcite . at an age of @xmath123myr ,
perhaps 100 sne have already occured in w1 ( see , e.g. , the discussion in * ? ? ?
the presence of an isolated x - ray pulsar confirms that supernovae have occurred there .
however , the likelihood of a recent snr contributing to the diffuse emission depends on the recent occurrence of a sn event in or near the core , as discussed in @xcite and @xcite .
the x - ray point sources from a _
observation are analyzed and reported by @xcite , while the diffuse emission is analyzed by @xcite .
the diffuse emission has an intrinsic ( @xmath118kev ) luminosity of @xmath124 , and a lorentzian spatial distribution with a hwhm along the major axis of @xmath125 arcseconds ( @xmath59pc ) , and a 5arcmin halo .
the emission ( in the energy range @xmath126kev ) can be fitted with a soft thermal component ( @xmath127kev ) , plus either a harder thermal component ( @xmath128kev ) with a low ( @xmath129solar ) fe abundance , or a nonthermal component with a power - law index @xmath130 .
the absorbing column , @xmath131 . in the thermal model ,
@xmath86 increases with distance from the cluster , while in the non - thermal model , @xmath19 is significantly steeper in the centre - most region considered .
there is no evidence for a recent sn explosion .
less than @xmath132 of the mechanical luminosity is dissipated as @xmath118kev x - rays so it is conjectured that a large fraction escapes into the ism .
however , the x - ray halo between @xmath62arcmin ( @xmath133pc radius ) is observed to attain a constant surface brightness of @xmath134 , which is not consistent with a cluster wind where almost all of the diffuse x - ray emission is produced within the core radius @xmath135 @xcite .
a thermal interpretation of the halo of diffuse emission is further challenged by the high temperature and lack of line emission .
more recently , @xcite analyze an xmm - newton pointing and determine that the hard component in an inner 2arcmin radius region is actually thermal , with a clearly detected he - like fe6.4kev line .
no evidence of a non - thermal component was found .
they report that the diffuse emission has a @xmath118kev luminosity of @xmath136 .
the lagoon nebula is an hii region associated with the young ( @xmath137myr ) open cluster ngc6530 , which contains several o - stars and about 60 b - stars .
it is about 1.3kpc away ( see * ? ? ?
* and references therein ) .
ongoing star formation occurs in several places , notably the hourglass nebula ( the brightest part of m8 ) and m8e .
the hourglass nebula is illuminated by an o7v star ( herschel36 ) .
@xcite argue that ngc6530 is slightly younger than the orion nebula cluster ( onc ) , being @xmath138myr assuming the onc is 2myr old .
if the onc is actually 3myr old , this would give the hourglass nebula an age of 2.5myr .
the lagoon nebula is summarized in @xcite .
@xcite claim that soft diffuse emission was `` probably '' detected from the southern lobe of the hourglass nebula .
the emission can be fitted with an absorbed mekal model with @xmath139 , @xmath140kev and an intrinsic ( @xmath141kev ) luminosity of @xmath142 .
however , there is unboutedly some contamination from unresolved point sources .
no diffuse emission is seen from a qualitative examination of a _
observation of m8 which did not cover the hourglass nebula @xcite - see also @xcite .
the arches cluster , like the quintuplet cluster , lies close to the galactic centre , being only 26pc away in projection ( see , e.g. * ? ? ?
it is slightly younger ( @xmath143myrs , * ? ? ?
* ; * ? ? ?
* ) and more massive @xcite than the quintuplet cluster .
@xcite have recently measured the kinematic mass of the cluster using _ keck_-lgs adaptive optics .
the deepest _ chandra _ observation to date is by @xcite . diffuse thermal x - ray emission with a prominent fe
k@xmath144 6.7kev emission line is seen from the core of the arches cluster .
the surface intensity declines steeply with radius , consistent with a cluster wind origin .
this central ( @xmath145pc ) `` plume '' region can be fitted with a single temperature nei plasma model with @xmath146kev , @xmath147 , @xmath148 and an intrinsic ( 2 - 8kev ) luminosity of @xmath149 .
in contrast , the emission in the outer regions of the cluster shows a prominent line at 6.4kev , a power - law continuum emission of non - thermal origin , and a non - axissymmetric spatial distribution with a bowshock morphology .
this may result from an ongoing collision between the cluster and the adjacent molecular cloud , which has a relative velocity of @xmath150 .
the interpretation of the 6.4kev fe k@xmath144 flourescence emission from neutral fe is still debated , with @xcite favouring a cosmic ray origin but @xcite favouring photoionization of nearby molecular clouds by x - ray photons .
@xcite find that the se extension ( which is where the 6.4kev emission is ) is best fitted with a pl+gaussian spectral model with @xmath151 and @xmath152 , and has an intrinsic ( @xmath118kev ) luminosity @xmath153 .
an even more extended `` lsbxe '' region is fitted with a mekal+pl+gaussian spectral model with @xmath154kev , @xmath155(fixed ) and @xmath156 , with an intrinsic ( @xmath118kev ) @xmath157 .
ngc3576 is a giant hii region located at a distance of @xmath158kpc @xcite , and which is projected to within 30 arcminutes of the more distant region ngc3603 .
together these regions make up the rcw57 complex .
the ionizing cluster for ngc3576 remains deeply embedded in the centre of an extended filamentary dust cloud , and not enough massive stars have yet been found to account for the radio luminosity @xcite .
@xcite analyzed two _ chandra _ pointings .
a southern pointing was centered on ngc3576 , while a northern pointing was designed to search for a young stellar cluster associated with the o8v+o8v eclipsing binary hd97484 ( emcar ) and the o9.5ib star hd97319 .
diffuse emission is seen to the se of ngc3576 ( hereafter identified as ngc3576s ) , while hard x - rays were seen in the northern pointing .
@xcite identified these sources as ngc3576s and ngc3576n , respectively .
the northern pointing revealed a young cluster ( termed ngc3576ob ) which appears older than ngc3576 to the south .
ngc3576s requires a 2-temperature spectral fit , with @xmath159 and @xmath160 .
the absorbing columns are @xmath161 and @xmath162 , respectively .
the softer component dominates the total emission which has an intrinsic @xmath163 .
@xcite suggests that the hot plasma responsible for this emission has forced itself out through a low - density pathway , analogous to the outflow seen from m17 , but seen more face - on and at a slightly earlier phase .
a gaussian at 0.72kev ( which accounts for 16% of the total emission ) is required for a good fit .
this may represent charge exchange processes .
in contrast ngc3576n requires the presence of a power - law component in spectral models .
the intrinsic luminosity is @xmath164 , 24% of which is contributed by a power - law continuum . a 3-temperature model is also required for a good fit , with the hardest nei component ( @xmath165kev ) accounting for 48% of the total emission .
@xcite speculate that the diffuse emission from this region has been enhanced by a recent cavity sn .
there is no evidence for charge exchange .
the luminous giant hii region ngc3603 contains the compact star cluster hd97950 , which is one of the most massive young star clusters in the milky way .
it contains 3 core h - burning wn - stars and up to 50 o - stars @xcite .
the most massive stars in the core appear to be coeval with an age of about 1myr , while less massive stars and stars in the cluster outskirts appear to be older ( * ? ? ?
* ; * ? ? ?
* and references therein ) .
it shows clear mass segregation , despite its young age .
@xcite suggest that dynamical processes may have been dominant for the high mass stars .
star formation appears to have occurred almost instantaneously , with @xcite deriving an upper limit to the age spread of 0.1myr .
the distance to ngc3603 is thought to be @xmath166kpc @xcite .
@xcite explore whether a phase of substantial gas - expulsion has occurred in ngc3603 .
a _ chandra _ cycle 1 observation was presented by @xcite , who noted diffuse x - ray emission within a central region of 2 arcmin radius with an intrinsic luminosity @xmath167 . however , this is 20% of the integrated point source emission within this region and may be completely due to undetected point sources .
@xcite recently re - analyzed this observation , finding 1328 point sources compared to the 348 sources found by @xcite .
the diffuse x - ray emission is anti - coincident with the mid - ir emission which traces the surrounding heated dust .
this is consistent with the hot plasma from the shocked stellar winds filling the cavities that they have carved . excluding an area around the core of ngc3603 ( which is likely dominated by unresolved point sources ) and a region to the west ( which may contain foreground emission related to the ngc3576 cluster ) , the diffuse x - ray emission is dominated by an nei thermal plasma component with @xmath168kev , @xmath169 , @xmath170 , and which contributes 86% of the total intrinsic @xmath171 .
no evidence for charge exchange processes was found though the exposure is quite short .
the flame nebula , ngc2024 , is one of the nearest sites of massive star formation ( @xmath172pc , * ? ? ?
it is part of the orionb giant molecular cloud ( e.g. * ? ? ?
* ) and is near the horsehead nebula .
a 3d structure of the region was proposed by @xcite ( see also * ? ? ?
@xcite suggested that the o8v - b2v star irs2b is the ionizing source of the hii region , but @xcite note that it could be a supergiant .
the age of ngc2024 is unclear , with estimates ranging from 0.3myr @xcite to several myr @xcite .
diffuse x - ray emission with a radius of 0.5pc from the centre of ngc2024 was reported by @xcite .
the emission has a very hard continuum ( @xmath173kev ) and shows a he - like fek@xmath144 line .
fitting the data with a `` leaky absorber '' model ( where emission from a single temperature plasma reaches the observer via two paths with different absorption ) returns @xmath174kev with @xmath175 and @xmath176 .
the intrinsic x - ray luminosity in the @xmath177kev band is @xmath178 .
@xcite note that a single massive star with a wind comparable to , or stronger than , that of a typical b0.5v star has enough energetics to power the observed x - ray emission .
this work shows that diffuse emission is present in a msfr in which only late o to early b stars exist .
k. a. , morse j. a. , kirshner r. p. , winkler p. f. , 2001 , in holt s. s. , hwang u. , eds , young supernova remnants vol .
565 of american institute of physics conference series , fabry - perot [ o iii ] @xmath1795007 observations of the smc oxygen - rich snr 1e 0102 - 72.9 .
pp 193196 u. , canizares c. r. , markert t. h. , hughes j. p. , 1992
, in american astronomical society meeting abstracts # 180 vol .
24 of bulletin of the american astronomical society , high - resolution x - ray spectroscopy of the supernova remnant n132d .
p. 790
a. f. j. , corcoran m. f. , stevens i. r. , skalkowski g. , marchenko s. v. , mcke a. , ptak a. , koribalski b. s. , brenneman l. , mushotzky r. , pittard j. m. , pollock a. m. t. , brandner w. , 2002 , apj , 573 , 191 l. , feigelson e. , montmerle t. , broos p. , chu y .- h . , garmire g. , getman k. , 2005 , in sjouwerman l. o. , dyer k. k. , eds , x - ray and radio connections parsec - scale x - ray flows in high - mass star - forming regions | the x - ray emission from a simulated massive stellar cluster is investigated .
the emission is calculated from a 3d hydrodynamical model which incorporates the mechanical feedback from the stellar winds of 3 o - stars embedded in a giant molecular cloud ( gmc ) clump containing 3240m@xmath0 of molecular material within a 4 pc radius .
a simple prescription for the evolution of the stars is used , with the first supernova explosion at t=4.4myrs .
we find that the presence of the gmc clump causes short - lived attenuation effects on the x - ray emission of the cluster .
however , once most of the material has been ablated away by the winds the remaining dense clumps do not have a noticable effect on the attenuation compared with the assumed interstellar medium ( ism ) column .
we determine the evolution of the cluster x - ray luminosity , l@xmath1 , and spectra , and generate synthetic images .
the intrinsic x - ray luminosity drops from nearly 10@xmath2ergss@xmath3 while the winds are ` bottled up ' , to a near constant value of 1.7@xmath4 between t=14myrs .
l@xmath1 reduces slightly during each star s red supergiant ( rsg ) stage due to the depressurization of the hot gas
. however , l@xmath1 increases to @xmath5 during each star s wolf - rayet ( wr ) stage .
the x - ray luminosity is enhanced by 2 - 3 orders of magnitude to @xmath6 for at least 4600 yrs after each supernova ( sn ) explosion , at which time the blast wave leaves the grid and the x - ray luminosity drops .
the x - ray luminosity of our simulation is generally considerably fainter than predicted from spherically - symmetric bubble models , due to the leakage of hot gas material through gaps in the outer shell .
this process reduces the pressure within our simulation and thus the x - ray emission .
however , the x - ray luminosities and temperatures which we obtain are comparable to similarly powerful massive young clusters .
# 1to 0pt#1 [ firstpage ] feedback hydrodynamics x - rays |
advances in algorithms and hardware have greatly increased the power and scope of monte carlo studies of critical phenomena .
cluster algorithms , first introduced by swendsen and wang @xcite , largely overcome the problems of critical slowing and thereby permit the study of very large systems using a reasonable number of monte carlo steps .
parallel machines provide platforms on which very large systems may be efficiently studied .
several groups @xcite have implemented cluster algorithms on parallel machines to obtain high precision results for the three - dimensional ising model .
the invaded cluster ( ic ) algorithm @xcite is a new type of cluster algorithm with features that make it very attractive for high precision studies of critical phenomena .
first , it has less critical slowing than other cluster algorithms .
indeed , here we provide strong evidence that for the three - dimensional ising model there is no critical slowing down .
second , the ic algorithm does not use an estimate of the critical temperature as an input ; instead , the critical point is sampled without any fine - tuning of parameters and an estimate of the critical temperature is obtained as an _ output _ of the simulation . on the other hand , the ic algorithm samples the ic ensemble rather than the canonical ensemble . although the two ensembles are believed to be equivalent in the infinite volume limit , the finite - size scaling properties of the ic ensemble are not yet well - understood .
here we present results that test a simple finite - size scaling hypothesis and determine the associated finite - size scaling exponent . in this paper
we describe an implementation of the invaded cluster algorithm for distributed memory spmd ( single program multiple data ) parallel machines . using this algorithm we studied two- and three - dimensional ising systems up to size 4096@xmath4 and 512@xmath5 , respectively .
there are several motivation for this project .
the first is to develop an efficient implementation of the ic method for parallel computation .
this is a non - trivial problem because the ic method appears to be quite sequential .
the second objective is to better understand the dynamics of the ic algorithm and the finite - size scaling properties of the ic ensemble .
these properties are interesting in their own right and they must be understood before the algorithm can be used for high - precision studies . finally , we wish to obtain accurate results for the critical temperature and exponents for the three - dimensional ising model . with relatively modest effort
we have been able to obtain accuracy comparable to recent studies in the measurement of @xmath6 and somewhat less accuracy for the critical temperature .
we argue that the ic method is likely to be the most efficient method for obtaining very high precision estimates for the critical temperature and the magnetic exponent for the ising model and related spin systems . in sec .
[ sec : ic ] we describe the invaded cluster method and discuss our parallel implementation . in sec .
[ sec : per ] we discuss the performance of the parallel ic algorithm running on cm-5 machines . in sec .
[ sec : num ] we present some high - precision numerical results for the three - dimensional ising model using the parallel ic algorithm .
conclusions are presented in sec .
[ sec : con ] .
in cluster monte carlo methods a spin system is sampled in both its spin representation and in an associated _ graphical representation_. the graphical representation is an interacting percolation model that is usually defined on the bonds of the lattice .
the graphical representation by itself contains all of the information relevant for the study of the spin system .
a single monte carlo ( mc ) step consists of two parts , a ` bond move ' and a ` spin move . ' during the bond move , the spins are frozen and the graphical configuration ( bond configuration ) is updated in accord with the current spin configuration . during the spin move , it is the opposite : the bonds are frozen and the spin configuration is updated . in a critical region , tremendous improvements in the efficiency of cluster methods over traditional local dynamics can arise if the graphical representation has a coinciding percolation threshold . in this case , clusters of spins on all scales are coherently updated . since the ic and swendsen - wang ( sw ) algorithms are very similar let us begin with a brief description of the sw algorithm .
consider the nearest neighbor ising magnet on the square or cubic lattice defined by the hamiltonian @xmath7 where @xmath8 , @xmath9 is the dimensionless coupling , and the sum is over all the bonds of the lattice . during the bond move , each ` satisfied ' bond ( @xmath10 is satisfied if @xmath11 ) is independently occupied with probability @xmath12 or left vacant with probability @xmath13 .
unsatisfied bonds are never occupied .
after a bond configuration has been created in this fashion , _ clusters _ are identified : a cluster is a set of ( like ) spins connected by occupied bonds ; an isolated spin is also counted as clusters .
next , the spin configuration is erased and the bond move is completed .
the spin move for the sw algorithm consists of assigning either spin up or spin down with probability 1/2 independently to every cluster .
each spin in the cluster takes the spin value assigned to the cluster and then the bond configuration is erased and the spin move is completed .
the sw algorithm samples the canonical ensemble at coupling @xmath9 where @xmath9 is related to @xmath12 via @xmath14 the spin move for the ic algorithm is the same as for the sw algorithm : each cluster is assigned either spin up or spin down with probability 1/2 .
the ic bond move consists of occupying the satisfied bonds in random order until the cluster configuration first fulfills a stopping rule . during the bond move
, bonds are occupied one at a time and after each bond is added , the cluster structure is checked to see if the stopping rule is satisfied .
there are a number of useful stopping rules , some of which , evidently , cause the algorithm to sample the ising critical point . in this study
, we will use ` topological ' rules which are appropriate for systems with periodic boundary conditions .
there are several versions : the ` 1-span ' topological rule demands that some cluster is connected around the system in at least one of the periodic directions as soon as this condition is fulfilled , the bond move is complete and no more bonds are added to the configuration . in general , the @xmath15-span rule ( @xmath16 ) requires that there is connectivity around the system in at least @xmath15 directions .
although all of the @xmath15-span rules sample the ising critical point , the different rules have different finite - size scaling behavior .
the critical coupling and the magnetic exponent may be obtained in the ic approach .
we begin with @xmath17 .
let @xmath18 be the ratio of the number bonds occupied during the bond move to the number of satisfied bonds for a single monte carlo ( mc ) step .
the average value of @xmath18 is an estimator of the critical temperature according to the relation @xmath19 where @xmath17 is the critical coupling and @xmath20 is the system size . to understand this relation , note that the only difference between the sw and ic algorithms is the way in which @xmath18 is determined in a bond move . for the sw algorithm , each satisfied bond
is independently occupied with probability @xmath12 so that the ratio of occupied to satisfied bonds is a binomial random variable . for the ic algorithm ,
the stopping rule determines the number of occupied bonds .
however , for large systems , one can presume that the fluctuations in @xmath18 become small for both algorithms . as a consequence , the ic algorithm that outputs @xmath21 yields the same value for all local observables as the sw algorithm with @xmath22 so that the relation @xmath23 yields an estimate of the temperature of the ic ensemble .
the bond percolation threshold on the set of satisfied bonds at the ising critical point is known to be @xmath24 .
since the @xmath15-span rules forces the cluster configuration to the percolation threshold , we see that eq .
( [ eq : fbeta ] ) holds . this argument ( which still lacks a rigorous proof ) is discussed in more detail in ref .
@xcite let us now turn to a discussion of the magnetic exponent .
for a critical stopping rule , such as a topological rule , the cluster that fulfills the stopping condition is called the _ spanning cluster_. this spanning cluster can be thought of as a typical example of a large scale cluster at criticality and thus its fractal dimension yields the magnetic exponent for the system .
in particular , if @xmath25 is the number of sites in the spanning cluster , we determine the fractal dimension @xmath26 via @xmath27 the fractal dimension is related to other critical exponents via @xmath28 where @xmath29 is the magnetic exponent , @xmath30 the magnetization exponent , @xmath31 the correlation length exponent and @xmath32 the spatial dimension . at the present time
we do not know how to extract the thermal exponent from the ic algorithm .
the finite - size scaling properties of the ic ensemble are currently under investigation . here
we state the fundamental assumption that will be used later in the analysis of the data .
let @xmath33 be the ( cumulative ) distribution function for @xmath18 measured in a system of size @xmath20 .
we suppose that there is an exponent , @xmath34 and function @xmath35 such that @xmath36\ ] ] naive finite - size scaling arguments suggest that @xmath37 with @xmath31 the ising correlation length exponent , however it is quite apparent that this is not the case .
measurements of @xmath34 reported in ref .
@xcite and below in sec .
[ sec : num ] yield a value @xmath38 in @xmath39 .
presumably this @xmath34 is the inverse of the correlation length exponent for a related percolation
ising hybrid model but at present , no definitive conclusions have been obtained .
the approach taken here to parallelize the ic algorithm is based upon the parallel sw algorithm described in ref .
the basic idea for both algorithms is to divide the lattice into cells each of which is handled by a single processor . during the bond move the cluster configuration
must be determined .
every site of the lattice has an integer label that identifies its cluster membership .
two sites are in the same cluster if they have the same label . at the beginning of the bond move
, every site has a unique local label .
the sw bond move proceeds in two stages . in the _ local _ stage ,
each processor in parallel occupies a fraction @xmath12 of the satisfied bonds in its cell and identifies the local cluster structure using a conventional sequential method .
the local clusters are the clusters that would exist if no bonds on the boundaries of the cell were occupied . during the _ relaxation _ stage ,
global labels are created and global clusters are identified .
global clusters which are formed out of local clusters by boundary connections between cells are assigned the lowest valued label from the the labels of the connected local clusters .
this is accomplished iteratively by the exchange of labels between neighboring cells .
a very similar relaxation stage is used for the ic algorithm except that more information must be exchanged between nodes to allow for checking the stopping condition . since
communication between processing nodes is much slower than processes handled by a single node it is desirable that individual nodes do as much work as possible between communication steps . on the other hand , the sequential ic algorithm checks for spanning after each new bond
is occupied .
clearly this would not be a good approach to directly parallelize .
instead , the algorithm converges to the value of @xmath18 by a sequence of choices of @xmath12 .
imagine that an independent uniform random number between 0 and 1 is assigned to each bond of the lattice .
( in fact , this is not quite the approach that is actually used but , for purposes of explanation , suppose that this is the case . ) on the first step each processor provisionally occupies all satisfied bonds with random numbers less than @xmath40 .
the resulting cluster configuration is identified and the stopping rule is checked .
for the sake of argument , suppose that spanning is not detected .
in this case , the provisionally occupied bonds are permanently occupied and a new value @xmath41 is chosen . the current cluster configuration is stored and additional bonds are provisionally occupied with random numbers from @xmath40 to @xmath42 . the cluster configuration is modified accordingly and the stopping condition is checked .
if spanning is not detected , the new cluster configuration is stored and the provisionally occupied bonds given tenure .
the value of @xmath12 is incremented in this way until on the @xmath43th step the stopping condition is fulfilled . at this point we know that spanning occurs between @xmath44 and @xmath45 .
we backtrack to the cluster configuration ( saved in memory ) associated with @xmath44 and try again , this time choosing @xmath46 .
henceforth , the algorithm carries out a binary search , each time reducing the range of @xmath12 by half until the stopping condition is just fulfilled and @xmath18 is determined . in the above discussion we supposed that random numbers are assigned to each bond and that on each step all satisfied bonds with random numbers less than @xmath12 are occupied
this is not an efficient procedure because it requires examining every bond each time @xmath12 is changed .
instead we do the following .
each processor creates a random permutation of the @xmath47 bonds in its cell .
this permutation determines the order in which bonds will be tested by the processor . on the first step , processor @xmath48
tests its first @xmath49 bonds where @xmath49 is chosen from the binomial distribution , @xmath50 $ ] for @xmath47 trials with success probability @xmath40 .
this is statistically equivalent to testing all bonds with random numbers less than @xmath40 as described in the previous paragraph .
suppose that spanning is not found , then @xmath40 is incremented to @xmath42 .
at this point processor @xmath48 has permanently occupied @xmath51 bonds .
processor @xmath48 now tests the next @xmath52 bonds where now @xmath52 is chosen from @xmath53 $ ] .
this process of occupying bonds continues as long as the stopping condition is not fulfilled and on the @xmath54th such step the number of occupied bonds is incremented according to @xmath55 .
suppose that spanning is first detected on step @xmath43 and that @xmath56 bonds have been provisionally occupied by processor @xmath48 during this step .
on the next step each processor tries to occupy a smaller number of bonds corresponding to splitting the difference between @xmath44 and @xmath45 . to do this , processor @xmath48 tests the next @xmath57 bonds starting with the @xmath58th bond where
@xmath57 is chosen from @xmath59 $ ] .
a little reflection shows that this procedure is statistically equivalent to the one described above for carrying out the binary search .
the advantage of this approach is that on successive steps it is not necessary to check every bond in the cell .
note that the local ordering of bonds is fully specified at the outset of the bond move but that the global ordering of the bonds is generated as needed and never fully specified .
we tested the parallel ic algorithm for the two - dimensional ising model on a 32-node cm-5 machine .
[ fig : fig1 ] shows that the number of relaxation cycles @xmath60 per mc step as a function of system size @xmath20 . the roughly logarithmic increase in @xmath60 results from the binary search method used to find @xmath18 .
we separately measured the time spent in local cluster growth @xmath61 and relaxation @xmath62 .
the local time increases as the number of spins per node , thus , for a fixed number of processors , @xmath63 .
the relaxation time is dominated by communication between nodes so @xmath62 is proportional to the boundary area of the local lattice and the number of relaxation cycles , @xmath64 .
[ fig : fig2 ] shows local and relaxation times as a function of system size . the relaxation time is greater for small systems but , when the number of spins per node exceeds several hundred thousand the local time dominates . in fig .
[ fig : fig3 ] we compare the performance of the parallel ic algorithm with that of the parallel sw algorithm for the two - dimensional ising model .
the total time per spin per mc step for both algorithms decreases as the system size increases and saturates for large systems . for large systems in both two and three dimensions ,
the time for one mc step for the parallel ic algorithm is about a factor of @xmath65 larger than for the parallel sw algorithm .
we obtained @xmath66s per spin per mc step for the parallel ic algorithm for the two - dimensional ising models with system size @xmath0 on the 32-node cm-5 machine and @xmath67s per spin per mc step for the three - dimensional ising model on the same machine and with the same number of spins ( @xmath68 ) .
the slower speed for three - dimensions results from the need for more communications between nodes since each node has six neighbors rather than four .
for both two and three dimensions the time per spin per mc step decreases as the system size increases .
the relaxation time is , however , still comparable to the local time for @xmath68 , so the parallel ic algorithm for the three - dimensional ising model can be expected to become more efficient as the system size increases holding the number of processors fixed . to test the sensitivity of the algorithm to the random number generator , we performed simulations using two different random number generators .
most of the data is collected using the 250-shift generator in which new random numbers are produced by @xmath69 , where @xmath70 are 32-bit integers and @xmath71 represents the bitwise exclusive or operation .
we also tested a combined shift register generator , introduced in ref .
the combined generator consists of two shift registers @xmath72 and @xmath73 with 32-bit integers @xmath74 .
new random numbers of the generator are combined by @xmath75 .
we observed no difference between results from the two generators .
for example , the mean value of @xmath18 for system size @xmath76 was measured as @xmath77 for the first generator and @xmath78 for the second .
we studied the three - dimensional ising model using the parallel ic algorithm for system sizes @xmath79 to @xmath1 . unless otherwise stated we used the 1-span rule . for each run
we allowed 200 mc steps for equilibration before collecting data .
independent runs consisted of @xmath80 mc steps for the smallest systems ( up to @xmath81 ) to @xmath82 mc steps for the largest system ( 512@xmath5 ) . for each size
, a number of independent runs were made .
we obtained statistics for @xmath18 and the size of the spanning cluster .
error bars represent the standard error ( one standard deviation ) associated with these independent runs .
the data for the 1-span rule is given in table i. figure [ fig : fig4 ] is a double logarithmic plot of @xmath83 vs. system size .
a fit of the form @xmath84 , yields @xmath85 and @xmath86 .
this result strongly supports the hypothesis that @xmath87 approaches zero as a power law . the finite - size scaling hypothesis , eq .
( [ eq : fss ] ) implies an asymptotically linear relation between @xmath21 and @xmath87 .
figure [ fig : fig8 ] shows plots of @xmath21 against @xmath87 for the three @xmath15-span rules . although these plots are consistent with asymptotic linear behavior , it is clear that there are large subleading corrections to scaling . for the 1-span rule
these corrections are negative and for the 3-span rule they are positive .
figure [ fig : fig5 ] shows more detail for the 1-span rule where , over the range of system sizes studied , @xmath21 varies only in the fifth significant digit .
however , to obtain yet higher accuracy than @xmath88 requires either larger system sizes or a good fit to the subleading terms .
fitting to the form @xmath89 we obtain @xmath90 with @xmath91 and @xmath92 .
the corresponding critical coupling , quoted with two standard deviation errors , is @xmath93 .
the fit is shown as a dotted line in the figure .
the primary difficulty in obtaining a high precision value of the critical coupling is the relatively large corrections to scaling , represented by the fact that the fitting parameter @xmath94 is a good deal larger than @xmath95 .
our estimate of the critical coupling is consistent with recent values @xcite ( e.g. @xmath96 @xcite ) but the error is much larger .
figure [ fig : median ] shows the median value of @xmath18 for the 1-span rule versus @xmath87 .
according to the finite - size scaling hypothesis this plot should also be linear and , indeed , approximately linear behavior is evidenced over the whole range of system sizes however it should be noted that the variation in the median is much larger than for the mean .
furthermore , errors in the mean are less . thus appears that the mean is a better estimator of the critical coupling .
we measured the mass ( the number of sites ) @xmath25 of the spanning clusters .
using eq .
( [ eq : d ] ) we can find the fractal dimension of the spanning cluster and the magnetic exponent . the double logarithmic plot of @xmath25 and @xmath20 in fig .
[ fig : fig7 ] shows that the data is well described by a power law and an examination of different ranges of system size yields no clear trends in the power . a linear fit for @xmath97 yields @xmath98 consistent with a recent value @xmath99 @xcite .
next we consider the dynamical properties of the algorithm .
the normalized autocorrelation function of an observable @xmath100 is defined by , @xmath101 where @xmath102 is time in monte carlo steps .
figure [ fig : fig9 ] shows the autocorrelation function for @xmath18 for a 128@xmath5 system .
the striking feature of this curve is that it is negative on time step 1 indicating anti - correlations between successive mc steps . as discussed in refs .
@xcite these anti - correlations are associated with the negative feedback mechanism that forces ic dynamics to the critical point . by the fourth step it is not possible to distinguish the autocorrelation function from zero .
the integrated autocorrelation time @xmath103 is defined by , @xmath104 @xmath105 using the procedure described in ref .
@xcite , we measured integrated autocorrelation times for the magnetization @xmath54 , energy @xmath106 and temperature estimator @xmath18 .
the autocorrelation times are given in table ii and plotted against system size in fig .
[ fig : fig6 ] .
due to these anti - correlations @xmath107 and @xmath108 _ decrease _ with system size ; @xmath109 remains nearly constant .
thus ic dynamics for the three - dimensional ising model apparently does not suffer critical slowing .
( although , as is the case for all cluster algorithms , the parallel time needed for a single mc step must increases logarithmically with the system size @xcite ) .
the dynamic exponent @xmath110 is defined by @xmath111 .
we obtained the dynamic exponents @xmath112 , @xmath113 and @xmath114 for @xmath54 , @xmath106 and @xmath18 respectively .
figure [ fig : fig10 ] shows @xmath115 as a function of @xmath116 .
it appears that the infinite volume limit for @xmath115 lies between between @xmath117 and @xmath118 . since @xmath119
there must be corresponding positive values summing to at least @xmath120 to @xmath121 for @xmath122
. it would be interesting to have more information about the infinite volume limit of the @xmath18 autocorrelation function .
the autocorrelation function is related to the standard error , @xmath123 in measuring @xmath124 according to , @xmath125 with @xmath126 and @xmath127 the number of monte carlo steps . in most monte carlo applications
the second term is much smaller than the first however for the ic algorithm this may not be the case since @xmath128 .
the autocorrelation function for @xmath18 is approximately @xmath129 for @xmath130 and small for @xmath131 so that the approximate error in measuring the critical coupling is @xmath132 suppose that the number of mc steps is large enough that @xmath133 as is the case here for the smaller system sizes . combining the scaling behavior of @xmath107 and of @xmath87 we have , for @xmath134 , @xmath135 on the other hand , if @xmath107 and @xmath136 are comparable as is the case here for the largest lattices then @xmath137 and @xmath138 .
we used a parallel implementation of the invaded cluster algorithm to study two- and three - dimensional ising models .
we have found that the parallel algorithm is quite efficient for large system sizes where most computational work is carried out constructing clusters within processing nodes and relatively little time is spent in communication .
we have shown that the invaded cluster algorithm has no critical slowing down for the three - dimensional ising model as measured by integrated autocorrelation times for the magnetization , energy or estimated critical temperature .
indeed , the integrated autocorrelation time for the latter two quantities approaches zero as the system size grows so that relatively few monte carlo steps are needed for high accuracy estimates of the finite - size critical energy or critical temperature .
on the other hand , extrapolation to the infinite volume limit for quantities such as the critical temperature is complicated by the absence of a well - understood finite - size scaling theory for the invaded cluster ensemble .
we have examined the hypothesis that the distribution of @xmath18 is controlled by a single scale factor and found that this apparently holds for large systems ( greater than @xmath139 ) but that there are significant corrections to the predicted asymptotic linear dependence of @xmath21 on @xmath87 .
this is not unexpected since the width of the @xmath18 distribution is rather large compared to @xmath140 .
thus relatively small corrections to scaling for the @xmath18 distribution can lead to relatively large deviations from the proposed linear behavior of @xmath21 vs. @xmath87 .
this means that to confidently reduce systematic errors in estimating the infinite volume critical temperature we would either have to go to much larger systems to make @xmath87 small or we would need a detailed understanding of the corrections to scaling in the ic ensemble .
our current estimates of the critical temperature using the ic method are consistent with the best conventional estimates but have errors which are about an order of magnitude larger .
in contrast to the critical temperature , corrections to scaling in measuring the critical dimension of the spanning cluster are very small and we are able to obtain an excellent estimate of the magnetic exponent . our result is @xmath141(@xmath3 ) .
a. n. burkitt and d. w. heermann .
54:210 , 1989 .
c. f. baillie and p. d. coddington .
b43:10617 , 1991.j .
kertesz and d. stauffer .
c3:1275 , 1992 .
j. apostolakis , p. coddington and e. marinari .
_ europhys .
_ 17:198 , 1992 .
r. hackl , h. -g .
mattutis , j. m. singer , th .
husslein and i. morgenstein .
_ physica _ a212:277 , 1994 . g. t. barkema and t. macfarland .
_ e50:1623 , 1994 . | a parallel version of the invaded cluster algorithm is described .
results from large scale ( up to @xmath0 and @xmath1 ) simulations of the ising model are reported .
no evidence of critical slowing down is found for the three - dimensional ising model .
the magnetic exponent is estimated to be @xmath2(@xmath3 ) for the three - dimensional ising model . |
this paper was motivated by recent investigations @xcite , in which a real possibility of the experimental measurement of the thermal conductivity of graphene was established and its extremely large value was confirmed .
the proposed and performed experiments were based there on the effect of the temperature shift of the raman g peak position due to the anharmonicity of graphene lattice @xcite . since that shift
may be measured with sufficient accuracy , the thermal conductivity of laser heated graphene may be restored using the heat equation with known sources and given geometrical parameters of the graphene flake .
bearing in mind the efficacy of existing experimental methods we consider in this paper the influence of moderate concentrations of different point defects ( carbon isotopes , substituting atoms and chemically adsorbed molecules ) on the thermal conductivity of graphene .
note that real and numerical experiments showed that graphene can be doped if required with boron , nitrogen @xcite , aluminium @xcite etc with concentrations reaching as much as several per cent without breaking the graphene planar structure @xcite .
the main task of this work is the elucidation of how the thermal conductivity of doped graphene depends on the temperature and mass of defects using as a tool classical methods of the quantum theory of solids . in doing
so we restrict ourselves by modeling all the substituting atoms as isotopic defects , assuming that the account of bond weakening may only give an additional reduction of lattice thermal conductivity @xcite .
this paper is organized as follows . in sections 2
we discuss the main mechanisms of the thermal resistance in graphene . in sections 3 and 4
we describe briefly the effect of phonon - phonon scattering and edge scattering as the dominant thermal resistance mechanisms in the case of ideal graphene . in section 5 the explicit expression for the phonon relaxation time determined by phonon - defect scattering
is deduced as a function of the quasi - momentum for each phonon branch . in the last section
we summarize the obtained results in the form of a fitting formula for the thermal conductivity of a graphene flake as function of temperature , flake size , mass and concentration of dopant .
the heat carriers in graphene are the phonons and conductive electrons .
the ideal graphene is a semi - metal with massless charge carriers .
as follows , electron and phonon contributions to the heat transport should be in the ratio @xmath0 where @xmath1 and @xmath2 are characteristic velocities of acoustic phonons and electrons ( holes ) in graphene .
due to the near 50-fold difference in velocities of electrons and phonons , the electron contribution to the ideal graphene total thermal conductivity appears to be less than 1% @xcite .
besides in semi - metals the number of carriers at low and room temperatures is small compared to that of lattice atoms .
since the phonon scattering on free carriers is proportional to their concentration , then the effect of phonon - electron scattering on the lattice heat transport is small comparing to that of phonon - phonon and phonon - defect scattering @xcite .
therefore further in this paper we will assume that the electronic subsystem counts very little in the heat transfer as a whole and can be ignored , the more so that the error of present measurements of graphene heat conductivity is @xmath3% .
with this assumption the finite value of the thermal conductivity of real graphene results from phonon scattering on other phonons , flake edges and point defects .
phonon - phonon scattering is conditioned only by the anharmonic contributions to the potential energy of inter - atomic interaction .
the quasi - momentum and energy conservation laws only allow phonon scattering processes in which at least three phonons are involved .
as the probability of collisions of four or more phonons is rather small , especially at low temperatures where the equilibrium number of phonons is not great , their contribution to the thermal resistance may be ignored .
remind that multi - phonon scattering processes are divided into normal processes ( n - processes ) with the strict conservation of momentum and so - called umklapp processes ( u - processes ) , for which the conservation of total quasi - momentum holds up to a non - zero reciprocal lattice vector .
the normal processes do not change the total heat flux , thus they drop out from possible mechanisms of thermal resistance , while the u - processes contribute , in an important manner , to the thermal resistance of semiconductors and isolators at high temperatures @xcite . on the other hand , with decreasing temperature the contribution of u - processes to the thermal resistance of graphene
is suppressed by effects of the defect - induced and edge phonon scattering .
considering phonon - defect scattering , we restrict ourselves to study only the impact of isotopic defects on the heat transfer in graphene . note that along with @xmath4 isotopes ( present in natural carbon at a concentration of @xmath5 ) the chemically adsorbed atoms and substitutional atoms ( like boron , nitrogen , aluminium etc ) can be to some extent treated as isotopic defects . in reality
, concentrations of all mentioned defects may range from zero up to several per cent .
however , as we show below , for the low atomic mass of carbon even a concentration of heavy `` isotopic '' defects @xmath6% may double the thermal resistance . at low temperatures the intensities of all mentioned processes rapidly decrease and the edge scattering is thus rendered the main mechanism limiting the heat transport .
taking account of the very high debye temperature of graphene and exponential drop in the probability of u - processes there is a temperature range within which only edge and defect scattering should be considered .
taking into account all the above mentioned contributions , the total relaxation time of a certain phonon state is given by @xmath7 where @xmath8 is the phonon wave vector and @xmath9 is the oscillation branch ; @xmath10 , @xmath11 and @xmath12 are the relaxation times determined by the phonon - phonon , phonon - defect and edge scattering , respectively .
if the temperature of graphen is well below its debye temperature ( the latter is near 2000k ) , then the phonon - phonon interaction can be considered as a rather small perturbation and thus different phonon states can be considered as independent .
thus the total lattice thermal conductivity @xmath13 can be represented as a sum over all possible phonon modes @xcite @xmath14 where @xmath8 runs the first brillouin zone , @xmath15 is the specific heat per normal mode , @xmath16 where @xmath9 numbers the phonon branches , @xmath17 is the frequency of the @xmath18-phonon , @xmath19 is the phonon group velocity and @xmath20 is defined by ( [ tautotal ] ) .
since our main task is to study the impact of defects on the thermal conductivity of ideal graphene , we will not scrutinize the contribution of the phonon - phonon u - processes using the same general expression as in @xcite ; however , with @xmath21 defined as described below ( [ eigen ] ) : @xmath22 where @xmath23 is the atomic mass of @xmath24 isotope and @xmath25 are conventional debye frequencies for three acoustic phonon modes ( @xmath26 , @xmath27 and @xmath28 rad / s ) . in order to find from ( [ kappasum ] ) certain values of @xmath13
we use the explicit expressions for @xmath17 in graphene obtained in @xcite on the basis of the following computationally simple harmonic nearest neighbor model . considering the total potential energy @xmath29 of ideal graphene as a sum of three components depending on different types of interaction between the nearest neighbors : - in - plane central forces with a force constant @xmath30 ;
- in - plane non - central forces with a force constant @xmath31 ; - out - of - plane non - central forces conditioned by @xmath32-electrons with a force constant @xmath33 + and applying bloch s theorem one can reduce within the framework of this model the problem of finding the sought frequencies @xmath21 to evaluation of eigenvalues of the ideal graphene dynamical matrix @xmath34 where @xmath35 @xmath36 in this way we find three acoustic and three optical branches @xmath21 of phonon spectra ( figure [ phonondisp ] ) . with notation @xmath37 , \\
f_1(\mathbf{k } ) = 12 ( j_1 ^ 2 + 2 x_0 ) + f_0(\mathbf{k } ) ( j_1 ^ 2 + 8 x_0 ) , \\
f_2(\mathbf{k } ) = \left\{9 j_1 ^ 2 + x_1 + 2 x_2 \cos{(k_1 a\ ! - \!k_2 a ) }
\left(\cos{k_1 a } + \cos{k_2 a}\right ) \right . +
\\ \qquad\qquad \left .
+ 2 \cos{k_1 a } \cos{k_2 a } \left[x_2\ ! + \ ! x_1 \cos{(k_1 a\ ! - \ !
k_2 a)}\right ] - 3 j_1 ^ 2 f_0(\mathbf{k } ) \right\}^\frac{1}{2 } , \end{array}\ ] ] the expressions for the eigenfrequencies @xmath21 take form : @xmath38^{1/2},\\ \omega_{la , ta , lo , to}(\mathbf{k } ) = \left [ \frac{3 ( j_1 + 2 j_2)}{2 m } \pm \frac{\sqrt{2}}{4 m } \sqrt{f_1(\mathbf{k } ) \pm 2 j_1 f_2(\mathbf{k } ) } \right]^{1/2}. \end{array}\ ] ] the values of the force constants @xmath39 were chosen in @xcite to fit in the experimental data on inelastic x - ray scattering in graphene @xcite .
it is noteworthy that the model with these @xmath40 reproduces raman experiment results and also gives appropriate values for sound velocities @xmath41[const ] in graphene ( see @xcite ) .
we remark that the given expression ( [ eigen ] ) for the out - of - plane ( bending ) mode @xmath42 is inconsistent in the limit @xmath43 with known dependence @xmath44 .
however the quasi - linear dependence of @xmath45 which follows from ( [ eigen ] ) occurs in graphene @xmath42-dispersion curves obtained in different ways already for @xmath46 or for @xmath47 mev @xcite .
this gives grounds for using the formula ( [ eigen ] ) for @xmath48 k. it is worth recalling that the out - of - plane phonon mode add little to the thermal transport in pure graphene because of its low group velocity and large gruneisen parameter @xcite . by our calculations
, the contribution of this mode to the total graphene thermal conductivity at room temperature does not exceed @xmath49 .
la ( lo ) , ta ( to ) and za ( zo ) are longitudinal , transversal and out - of - plane acoustical ( optical ) branches respectively . ] in all our calculations we used the exact expressions ( [ eigen ] ) without any further simplification . at low temperatures ( @xmath50k )
this does not improve visibly the debye approximation .
however , the discrepancy between results of the debye approximation and those obtained using ( [ eigen ] ) are quite visible even at room temperature , especially for small graphene flakes ( figure [ termconddebye ] ) . )
( solid lines ) and obtained within the debye approximation ( dashed lines ) . ]
in the low temperature limit where the u - scattering processes are not active and defect scattering is relatively weak the phonon mean free path @xmath51 grows rapidly with @xmath52 . on the other hand @xmath51 can not exceed the lattice linear size .
thus the thermal conductivity in this region is limited only by the edge scattering and should depend linearly on the scaling factor for similar graphene flakes . for a bounded convex graphene flake with diameter much greater than the lattice constant ,
the distribution of eigenfrequencies is just the same as that in the case of infinite ideal graphene .
the phonon eigenfrequencies of the flake numbered in increasing order are well within the dispersion curves of ideal graphene .
though the displacement eigenvector @xmath18 for the infinite ideal graphene lattice is not an eigenvector for a bounded graphene flake , it is , however , the superposition of the flake s displacement eigenvector corresponding to frequencies close to @xmath17 . therefore the uncertainty of the exact eigenfrequencies can be reduced to the uncertainty of quasi - momentum @xmath8 : @xmath53 , where @xmath54 , @xmath55 is the flake s area and @xmath56 is its characteristic size . for the phonon wave associated with @xmath18 the energy - time uncertainty principle @xmath57 gives @xmath58 setting @xmath59 where @xmath60 is the length of the segment inside the flake of ray coming from the point ( x , y ) and @xmath61-inclined toward the @xmath62-axis , and @xmath63 is the dimensionless flake form - factor ( @xmath64 for the disc and reduces with increasing of the flake diameter ) we use further for the @xmath65 an approximate formula @xmath66 in which the ballistic mean free path @xmath67 absorbs all the peculiarities of the flake geometry . in the case of graphene with abundant linear structural defects
the boundaries of emerging domains became the main source of the edge scattering .
if so , to obtain @xmath68 one may calculate it as above for a certain domain with some shape and area @xmath69 and then average the obtained value over the distribution of the domain shapes and areas and also incorporating in this case the specularity parameters @xmath70 at the domain boundaries as in @xcite .
note that at low temperatures the finite size of a graphene flake is a dominant factor dictating the finiteness of the phonon free path .
since all the active phonons are acoustic phonons near the @xmath71-point where @xmath72 , then the thermal conductivity takes the form @xmath73 , where @xmath9 numbers acoustic phonon branches and @xmath74 are corresponding partial lattice specific heats per unit volume .
the mean free path of the @xmath18-phonon caused only by its scattering on point defects is defined by the evident expression @xmath75 where @xmath76 is the surface density of point defects and @xmath77 is the corresponding phonon total cross - section .
hence @xmath78 using the optical theorem and lippmann - schwinger formula for calculation of @xmath77 in ( [ taudef ] ) yields @xcite ( see chapter 11 ) @xmath79 where @xmath80 is the area of a two - dimensional graphene lattice , @xmath81 is the green s function of the infinite ideal graphene lattice with the dynamical matrix @xmath82 defined above ( [ dynamicalmatrix ] ) , and @xmath83 is the t - matrix , which connects the green s function @xmath84 of a defect lattice with that of the ideal lattice : @xmath85 , or @xmath86 . for an isotopic defect located at the lattice site ( 0,0 ) the element of t - matrix with indices @xmath87 ( where @xmath88 numbers graphene sublattices ,
@xmath89 is associated with spatial components of the atom displacement and @xmath90 is the lattice vectors ) has the form @xmath91 and @xmath92^{-1}\right)^{i , j}\mathbf{e}^{j}_{0 , \nu}(\mathbf{k } ) .
\end{array}\ ] ] the expressions ( [ taudefim ] ) , ( [ tandqmatrix ] ) , ( [ impart ] ) were immediately used for the calculation of @xmath93 .
we used directly the relations for the total lattice thermal conductivity ( [ tautotal ] ) , ( [ kappasum ] ) together with the explicit expressions discussed above ( [ tauu ] ) , ( [ tauedge ] ) , ( [ taudefim ] ) and ( [ eigen ] ) for @xmath94 and @xmath17 , to calculate the thermal conductivity of graphene flakes in a wide range of linear sizes , temperatures and concentrations of isotopic defects with @xmath95 varying from @xmath96 to @xmath97 .
the obtained data on graphene thermal conductivity in the range @xmath98k , @xmath99 , @xmath100 and @xmath101 are reproduced with an accuracy to @xmath102 by the following empirical formula which permits us to obtain the value of the thermal conductivity without re - calculating explicit expressions ( [ tauu ] ) , ( [ tauedge ] ) , ( [ taudefim ] ) and ( [ eigen ] ) ( any further improvement of this formula has no sense in view of the approximations laid down into the above theoretical constructions ) @xmath103 where @xmath104 is the thermal conductivity of an ideal graphene flake , @xmath105 , @xmath106 and @xmath107 is the characteristic length of the flake in micrometers , @xmath108^{3}+a_3 ^ 2(\mathcal{l})},\ ] ] @xmath109 and @xmath110 @xmath111 @xmath112 as expected , the results for ideal graphene are close to those of balandin et al .
notice that the qualitative trend of the thermal conductivity of doped graphene as a function of temperature , type and concentration of defects emerging from ( [ kappaapprox ] ) agrees well with arguments and qualitative conclusions for graphene of the comprehensive review paper @xcite . at low temperatures the dominant factor of phonon scattering is the finiteness of a graphene flake that gives @xmath113 .
for relatively high temperatures the thermal conductivity @xmath13 varies in inverse proportion to @xmath114 due to the phonon - phonon scattering .
the phonon - phonon scattering is calculated here by summation over all the brillouin zone in contrast to @xcite , where the debye approximation was used .
the debye approximation tends to increase the value of thermal conductivity since it gives non - zero sound velocities at the brillouin zone edges .
the difference between these approaches is noticeable for @xmath115k being as much as 20% for @xmath116k , but so far remaining within the experimental error . in order to avoid any additional geometrical and non - geometric parameters ,
the phonon mean free path associated with the edge scattering is taken simply as ( [ formfactor ] ) .
the established decrease of graphene thermal conductivity caused by isotopic defects appears to be approximately linear in concentration @xmath76 ( for small concentrations ) and strongly dependent on @xmath117 and temperature .
note that though the cross - section of phonon scattering on defects is by itself temperature independent , it depends substantially on the phonon quasi - momentum .
therefore the relative contributions of phonons from different parts of the first brillouin zone to the heat transport changes with temperature , resulting in the complex temperature dependence of the contribution of defects to the total heat resistance . for substitution defects with @xmath118 , such as boron , nitrogen and carbon isotopes
the relative reduction in thermal conductivity is just proportional to @xmath119 and is well within experimental error .
for example , for thermal conductivity of graphene with concentration @xmath120 of @xmath121c according to the above results at room temperature the derivative @xmath122 and varies from -1.2 to -0.8 as @xmath123 increases from 1 to 100 @xmath117 m . for the isotopically modified @xmath121c - graphene with concentration @xmath124 of @xmath125c
this derivative is just the same while the difference in heat conductivities of pure @xmath125c- and @xmath121c - graphene appears to be lesser than that between pure @xmath125c- and pristine graphene .
this agrees with experimental results on the thermal conductivity of graphene with concentrations of @xmath121c isotope of 0.011 and 0.992 @xcite . furthermore , despite the fact that the model developed above is not valid without further improvement for very high defect concentrations , the ratio of the calculated value of thermal conductivity of pristine graphene to that of 50 - 50 isotope mixture is the same as in @xcite . in the meantime we can not give so far the satisfactory explanation of the drastic drop in the measured value of graphene thermal conductivity for the @xmath126 isotope concentrations observed in @xcite . for heavier defects ( aluminium atoms and different chemically adsorbed molecules ) with @xmath127 the thermal conductivity drops sufficiently even for @xmath128 .
as an example , 1% of `` isotopic '' @xmath129al ( @xmath130 ) reduces the thermal conductivity by more than half ( figure [ termcond ] ) . strictly speaking ,
the replacement atoms and adsorbed molecules are not isotopic defects .
however , the entailed bonds weakening may only enhance the effect of low frequency phonon scattering ( see @xcite , chapter vi ) and thus lead to a further reduction in the thermal conductivity of graphene .
it should be noted that chemisorbed molecules may desorb in the temperature range 300 - 600 k decreasing the concentration of defects .
so the graphene thermal conductivity under gas - surface equilibrium conditions in corresponding temperature regions may show a sufficient rise with @xmath114 instead of the `` normal '' behavior @xmath131 .
thus the insertion of rather low concentrations of rather heavy dopants has a disproportionately great impact on the thermal resistance of graphene .
this conclusion is completely in line with results of rather recent molecular dynamics simulations of phonon transport in doped graphene @xcite .
the authors are grateful to the reviewers for their useful remarks and comments .
this work was supported by the ukrainian state fund for fundamental researches , grant no . 0112u001739 .
35 balandin a a , ghosh s , bao w , calizo i , teweldebrhan d , miao f and lau c n 2008 superior thermal conductivity of single - layer graphene _
nano letters _ * 8 * 902 - 7 .
ghosh s , calizo i , teweldebrhan d , pokatilov e p , nika d l , balandin a a , bao w , miao f and lau c n 2008 extremely high thermal conductivity of graphene : prospects for thermal management applications in nanoelectronic circuits _ appl .
lett . _ * 92 * 151911 .
nika d l , ghosh s , pokatilov e p and balandin a a 2009 lattice thermal conductivity of graphene flakes : comparison with bulk graphite _ appl .
* 94 * 203103 .
balandin a a , ghosh s , nika d l and pokatilov e p 2010 thermal conduction in suspended graphene layers _ fullerenes , nanotubes , and carbon nanostructures _ * 18 * 474 - 86 .
murali r , yang y , brenner k , beck t and meindl j d 2009 breakdown current density of graphene nanoribbons it appl .
. lett . * 94 * 243114 - 1 - 3 .
jauregui l a , yue y , sidorov a , hu j , yu q , lopez j , jalilian r , benjaming d k , delkd d a , wu w et al .
2010 thermal transport in graphene nanostructures : experiments and simulations _ ecs transactions _ * 28 * 73 - 83 .
chen s , moore a l , cai w , suk j w , an j , mishra c at al . 2011
raman measurements of thermal transport in suspended monolayer graphene of variable sizes in vacuum and gaseous environments _ acs nano _ * 5 * 321 - 8 .
chen s , wu q , mishra c , kang j , zhang h , cho k , cai w , balandin a a and ruoff r s. 2012 thermal conductivity of isotopically modified graphene _ nature materials _ * 11 * 203 - 7 .
nika d l and balandin a a 2012 _ j.phys . : condens
. matter _ * 24 * 233203 .
calizo i , miao f , bao w , lau cn and balandin a a 2007 variable temperature raman microscopy as a nanometrology tool for graphene layers and graphene - based devices _ appl .
lett . _ * 91 * 071913(1 - 3 ) .
calizo i , balandin a a , bao w , miao f and lau c n 2007 temperature dependence of the raman spectra of graphene and graphene multilayers _ nano letters _ * 7 * 2645 - 9 . li x , wang h , robinson j t , sanchez h , diankov g and dai h 2009 simultaneous nitrogen doping and reduction of graphene oxide
soc . _ * 131 * 15939 - 44 . dos santos m c and alvarez f 1998 nitrogen substitution of carbon in graphite : structure evolution toward molecular forms b
* 58 * 13918 - 24 .
ao z m , jiang q , zhang r q , tan t t and li s 2009 al doped graphene : a promising material for hydrogen storage at room temperature * 105 * 074307 .
berciaud s , ryu s , brus l e and heinz t f 2009 probing the intrinsic properties of exfoliated graphene : raman spectroscopy of free - standing monolayers _ nano lett .
_ * 9 * 346 - 52
. casiraghi c , pisana s , novoselov k s , geim a k and ferrari a c 2007 raman fingerprint of charged impurities in graphene _ appl .
lett . _ * 91 * 233108 - 10 .
ziman j m 1960 _ electrons and phonons . the theory of transport phenomena in solids _ ( oxford : clarendon press ) .
klemens p g 2001 theory of thermal conduction in thin ceramic films _ international journal of thermophysics _ * 22 * 265 - 75 .
nika d l , pokatilov e p , askerov a s and balandin a a 2009 phonon thermal conduction in graphene : role of umklapp and edge roughness scattering b * 78 * 155413 .
adamyan v and zavalniuk v 2010 phonons in graphene with point defects .
_ j.phys . : condens
. matter _ * 23 * 015402 - 1 - 10 .
mohr m , maultzsch j , dobardi e , reich s , miloevi i , damnjanovi m , bosak a , krisch m and thomsen c 2007 phonon dispersion of graphite by inelastic x - ray scattering b * 76 * 035439 .
oshima c , aizawa t , souda r , ishizawa y and sumiyoshi y 1988 surface phonon dispersion curves of graphite ( 0 0 0 l ) over the entire energy region * 65 * 1601 - 4 .
siebentritt s , pues r , rieder k - h and shikin a m 1997 surface phonon dispersion in graphite and in a lanthanum graphite intercalation compound b * 55 * 7927 - 34 .
wirtz l and rubio a 2004 the phonon dispersion of graphite revisited _ solid state communications _ * 131 * 141 - 52 .
politano a , marino a r , campi d , faras d , miranda r and chiarello g 2012 elastic properties of a macroscopic graphene sample from phonon dispersion measurements _ carbon _ * 50 * 4903 - 10 .
mounet n and marzari n 2005 first - principles determination of the structural , vibrational and thermodynamic properties of diamond , graphite , and derivatives b * 71 * 205214 . stoneham a m 1975 _ theory of defects in solids .
( oxford : clarendon press ) .
zhang g and li b 2010 impacts of doping on thermal and thermoelectric properties of nanomaterials _ nanoscale _ * 2 * , 1058 - 68 .
hao f , fang d , xu z 2011 mechanical and thermal transport properties of graphene with defects _ appl .
* 99 * 041901 .
zhang h , lee g , cho k 2011 thermal transport in graphene and effects of vacancies b * 84 * 115460 . | the thermal conductivity of doped graphene flake of finite size is investigated with emphasis on the influence of mass of substituting atoms on this property .
it is shown that the graphene doping by small concentrations of relatively heavy atoms results in a disproportionately impressive drop of lattice thermal conductivity . |
this report is based on a talk , presented at dis2000 in liverpool , with the purpose to summarize shortly the legacy of hera during the first decade of its operation .
the talk was given to an audience working in the field .
for the benefit of the reader who is less familiar with the subject , the proceedings version has been extended to include an appendix with definitions of the kinematical variables used in the text .
the main motivation for building the hera @xmath0 collider was to continue and study the inner structure of matter and the nature of forces in a way similar to that of rutherford , by probing the proton with higher and higher gauge - boson virtuality , @xmath1 , in order to resolve smaller and smaller distances .
this way of probing the proton is known as deep inelastic scattering ( dis ) .
the plane showing the fraction of the proton s momentum carried by the probed parton , @xmath2 , and the virtuality @xmath1 of the probing photon , is shown in figure [ fig : x - q2 ] and is impressive .
it shows how hera has extended the kinematic reach in both direction in @xmath1 , low and high , and in the low @xmath2 region , by few orders of magnitude .
before this talk , i polled the view of several eminent people as to what they consider to be the highlights of hera so far .
all of them unanimously put as number one the surprising sharp rise of @xmath3 with decreasing @xmath2 @xcite , an example of which is shown in figure [ fig : f2-q2 - 15 ] .
the dglap evolution equations predict the rise of the parton distributions with decreasing @xmath2 .
why is then the observed rise so surprising ?
the physical quantity which is measured is @xmath3 . to decompose @xmath3 into quarks , we need the ` impulse approximation ' to be valid .
this means that the configuration of the fock states into which the proton fluctuates has to be frozen during the interaction time .
this condition leads @xcite to a relation between the @xmath1 of the probing photon and some other scales like quark masses , @xmath4 , quark transverse momentum @xmath5 s and @xmath2 .
the fluctuation time , @xmath6 , is inversely proportional to the mass squared of the probed configuration , @xmath7 .
the interaction time , @xmath8 , is inversely proportional to the photon virtuality , @xmath9 .
the freezing condition for the use of impulse approximation requires @xmath10 or , @xmath11 it is therefore highly non trivial that at very low @xmath2 the expectations of the dglap evolution equations , which apply to partons , remain valid for @xmath3 .
however , hera teaches us that at @xmath12 the impulse approximation may be valid down to @xmath13 1.5 gev@xmath14 . as a summary of this point
i would like to quote peter landshoff : _ the amazing discovery of hera is the rapid rise of @xmath3 at small @xmath2 .
this remains a challenge for theory : there is no respectable theoretical understanding . _
the second subject , unanimously acclaimed as highlight , is the observation of abundant diffraction in dis @xcite .
this came as a surprise because if @xmath3 is an incoherent sum of quarks , a dis process on a quark can not be diffractive because of color .
a diffractive dis reaction could happen only if the fock state of the proton at the origin includes colorless objects .
if this were indeed the case , one would expect that diffractive dis processes would be soft in nature , contrary to the accumulated evidence .
one of the most striking behaviour of these large rapidity gap events is that they have a similar energy behaviour as that of the total inclusive cross section @xcite , as shown in figure [ fig : diff / tot ] , not expected in soft processes .
does that mean that these diffractive dis processes are all hard ?
when one thinks about a dis process involving partons , the first intuition is to think of it as a hard process .
this however turns out not to be the case .
the best way to understand this is when the interaction is viewed in the proton rest frame .
the photon fluctuates into a @xmath15 pair , which can either be a large spatial configuration , or a small one .
the aligned jet model of bjorken @xcite , leading to scaling properties of @xmath3 , assumes that only large configurations contribute to dis cross section .
that would make dis a predominantly soft process .
the small configurations lead to scaling violation , as expected in qcd . to the extend to which scaling violation introduces only logarithmic corrections to scaling ,
one is led to conclude that , even within qcd , dis is indeed predominantly soft . to me , it came as a cultural shocks to learn that dis does not necessarily mean a hard process . to confirm the validity of the picture above
, we would like to isolate events in which we know that the photon predominantly fluctuates into a small configuration .
this is achieved either by studying events initiated by longitudinal photons ( @xmath16 ) and looking at the behaviour of @xmath17 , or by studying exclusive vector meson electroproduction and deeply virtual compton scattering ( dvcs ) .
since , unfortunately , we did not measure the ratio @xmath18 of the longitudinal to transverse cross sections at hera , i used as an exercise the parameterization of @xmath19 by badelek , kwiecinski and stasto @xcite together with that of allm97 @xcite as a good representation of @xmath20 , to calculate the effective @xmath21 energy dependence of the cross section of @xmath17 and @xmath22 separately , assuming a @xmath23 dependence .
the result is shown in figure [ fig : sigl ] .
as one sees , @xmath17 has a very steep rise with @xmath1 , as expected from a small configuration process , reaching the value of 0.5 at @xmath13 1000 gev@xmath14 .
the energy dependence of the cross section for photoproduction of vector mesons @xcite , shown in figure [ fig : xsect - vm ] , indeed confirmed that when the photon is squeezed into a small size , there is a definite change in the @xmath24 behaviour , as expected from pqcd , which connects the @xmath24 behaviour with the rise of the gluons . in the case of @xmath25
this happens already at @xmath1=0 , due to the presence of the charm quarks .
for the lighter vector mesons , it happens around @xmath1=10 gev@xmath14 , as seen from the @xmath26 ratio @xcite , shown in figure [ fig : phi / rho ] , and which reaches the expected value of 2/9 at that @xmath1 .
the other property of hard exclusive processes is the universal slope of the diffractive peak . in regge language
this means a decreasing slope @xmath27 of the regge trajectory .
this is shown in figure [ fig : a - vector ] where the pomeron trajectory is presented as obtained from elastic photoproduction of @xmath28 , @xmath29 and @xmath25 @xcite .
one sees a clear flattening in case of the @xmath25 .
the inclusive diffraction structure function @xcite , presented in figure [ fig : diff ] shows a very different behaviour from that of the proton structure function @xcite , shown in figure [ fig : sc - viol ] .
while the proton structure function shows positive scaling violation at low @xmath2 and negative one at high @xmath2 , the diffractive structure function shows practically only positive scaling violation , and may reach scaling at a large values of @xmath30 .
qcd factorization was proven to hold also in the inclusive diffraction case and thus dglap evolution equations could be applied .
however it turned out that for the evolution equations to be able to describe the @xmath1 behaviour of the data , a large diffractive gluon component is required @xcite .
this brings me to talk about the gluon in the proton . what have we learned about the gluon ?
from the scaling violation , using the impressive measurements at hera , shown in figure [ fig : sc - viol ] , the precision of the gluon determination has improved tremendously , as shown in figure [ fig : gluon - h1 ] @xcite .
the direct gluon determination from charm production is in very good agreement with the indirect ones from scaling violation . however , when trying to stretch the evolution equations down to low @xmath1 ,
as presented in figure [ fig : gluon - z ] for @xmath1 = 1 gev@xmath14 , one finds a strange behaviour of the gluon at low @xmath2 @xcite .
what is this strange behaviour of the gluons at low @xmath2 and low @xmath1 tell us ? if anything strange is happening , the gluons will feel it first because of the 9/4 color factor : @xmath31 .
one can calculate the probability @xmath32 that an interaction on the gluon is associated with diffraction .
for a value of @xmath1=4.5gev@xmath14 and @xmath33 this comes out to be @xmath340.4 @xcite , which is very close to the black disc limit of 0.5 .
does this mean that we see some kind of saturation ( unitarity effects ) ?
such saturation effects could lead to the suppression of large @xmath15 configuration and lead to the breakdown of dglap evolution in this region .
this can in fact be nicely observed when @xmath3 is plotted as function of @xmath1 for fixed @xmath35 @xcite , displayed in figure [ fig : f2-lowq2 ] , where one sees the approximate scaling down to @xmath13 2 gev@xmath14 and then the @xmath36 decrease of @xmath3 towards the low @xmath1 region .
this behaviour , and also that of the caldwell plot @xcite , shown in figure [ fig : caldwell ] , are well described by the saturation model of golec - biernat and wuesthoff @xcite .
this section can be summarized by a quotation from al mueller s message : _ the turnover at small @xmath1 ( and small @xmath2 ) along with the success of the golec - biernat wuesthoff model in describing that turnover ( along with diffractive scattering ) suggest that unitarity limits have been reached when @xmath37 1 - 2 gev@xmath14 in the low @xmath2 region at hera . in parton language
this is saturation in the small @xmath2 wave function of the proton . _
hera , the machine designed to teach us about the proton , turned out to be a good tool for studying the structure of both the real and the virtual photon .
a clear two component structure of the real photon has been observed at hera @xcite .
the gluon distribution in the photon has been extracted and also shows a rise with decreasing @xmath2 @xcite .
hera thus provides an alternative way to that of the @xmath38 collider to measure the photon structure function and to extract parton distributions in the photon . studies of these type in the dis regime showed that also a virtual photon can have a two component structure at low @xmath2 even at virtualities as high as @xmath13 50 gev@xmath14 @xcite .
this of course raises the question : is the virtual photon probing the structure of the proton or are the partons from the proton probing the structure of the photon ? who is probing whom ?
in fact , in the proton rest frame @xmath2 can be given a spatial meaning : the coherence length can be written as @xmath39 which in the low @xmath2 region can be very large , much larger than the radius of the proton .
thus at low @xmath2 we are actually not studying the structure of the proton . to put it in bjorken s words :
_ the important spacetime regions are neither in the valence region of the hadron nor in the `` valence '' region of the photon probed . _ at low @xmath2
we are probing the structure of the vacuum .
the subject of the very large @xmath1 region @xcite actually belongs to the future legacy of hera , which the hera upgrade will produce .
so far it produced a text - book result showing that the cross section of nc and cc reactions meet at @xmath40 .
recently , the higher statistics data also showed clearly the effect of the electroweak interference and the first measurements of @xmath41 were obtained .
searches for deviation from the sm caused excitement .
limits on exotic processes were obtained and await higher luminosity data .
our discussion in the previous section clearly indicates that in order to study the interior of the proton , one needs not only to go to higher values of @xmath1 but also to @xmath2 values in excess of 0.1 .
the hera upgrade should provide the luminosity necessary for obtaining high enough statistics in the high @xmath1 high @xmath2 region .
the sharp rise of @xmath3 with decreasing @xmath2 , the diffraction in dis , should we have been surprised ?
some say yes , some say - in retrospect- no . the fact , however , is that we were surprised and unprepared .
the original design of the detectors was not really well suited for the low @xmath2 , low @xmath1 physics .
none of the dis mc generators had diffractive processes .
we also took seriously parton distributions which predicted a flat @xmath3 at low @xmath2 . to summarize the lesson learned from the low @xmath2 physics at hera
, we have observed a tremendous progress in understanding of qcd dynamics for high energy interactions and we might be faced with the existence of a new qcd regime with high parton densities .
another legacy of hera was pointed out by james stirling : _ studying qcd at hera has given us great confidence that we can very precisely predict lhc cross sections .
indeed , it is under active consideration to use standard model @xmath42 cross sections to measure the lhc luminosity , and it is only because the proton structure is pinned down so well by hera that we can trust the theoretical calculations . _ hera has actually elevated qcd almost to the status of qed , in the sense that qed processes are used to measure luminosities at hera . not that there were doubts in the correctness of qcd .
however , qcd is a complicated theory , and thus we can say that hera in the year 2000 is a glorious triumph of minds dealing with qcd .
did we however get a better picture of the proton ?
rutherford s experiment allowed him to say that most of the atom is empty with a nucleus at its center .
how do we see the proton ? some describe the proton as being a thompson like proton : soup of gluons filled with pointlike quark raisins. other see it as a rutherford like proton : 3 centers of pointlike parton clouds . asking lonya frankfurt ,
his reply was that the outer 20% of the proton is a pion cloud , while nothing is known about the inner 80% .
maria krawczyk brought to my attention the view of altarelli , cabbibo , maiani and petronzio @xcite , who view the proton as 3 constituent quarks , each of which is a complex object made out of pointlike partons . where are all these partons located in the proton ?
it would be nice to be able to stand in front of a class and tell the students , just as rutherford did about the atom , how the proton looks like .
the hera upgrade program will hopefully give us the answer .
i would like to thank the organizers of this workshop , headed by john dainton and josephine zilberkweit , who did a marvelous job .
thanks also for the lord mayor of liverpool , councillor edwin clein , who promised us that in liverpool all our dreams come true , and to the member of parliament for the liverpool riverside constituency , mrs .
louise ellman , for her understanding and support of basic research .
the discussion and input of the following people to this talk is highly appreciated : h. abramowicz , g. altarelli , j. bjorken , a. caldwell , j. dainton , m. derrick , r. devenish , r. eichler , f. eisele , l. frankfurt , d. haidt , r. klanner , m. krawczyk , p. landshoff , e. levin , r. nania , j. stirling , m. strikman , a. wagner , g. wolf , a. wroblewski .
this work was partially supported by the german - israel foundation ( gif ) , by the us - israel binational foundation ( bsf ) , and by the israel science foundation ( isf ) .
99 h1 collab . , i. abt et al . , ; zeus collab .
, m. derrick et al . , .
r. p. feynman , _ photon - hadron interactions _
, w. a. benjamin , inc .
( 1972 ) ; e. leader and e. predazzi , _ an introduction to gauge theories and modern physics _ , cambridge university press ( 1996 ) , vol . 1 , p. 382 .
zeus collab . , j. breitweg et al . , .
h1 collab .
, c. adloff et al . , .
j. d. bjorken , proceedings of the international symposium on electron and photon interactions at high energies , cornell university , 1971 .
b. badelek , j. kwiecinski and a. stasto , .
h. abramowicz and a. levy , desy 97 - 251 , [ hep - ph/9712415 ] .
for a recent review , see e.g. h. abramowicz , lp99 , stanford , desy 00 - 025 , [ hep - ph/0001054 ] . for a recent review , see e.g. m. klein , lp99 , stanford , [ hep - ex/0001059 ] .
zeus collab . , j. breitweg et al . , .
l. frankfurt and m. strikman , .
zeus collab . ,
j. breitweg et al .
, desy 00 - 071 , [ hep - ex/0005018 ] .
h. abramowicz and a. caldwell , .
k. golec - biernat and m. wusthoff , .
zeus collab . , j. breitweg et al . , .
h1 collab .
, c. adloff et al . , .
h1 collab . , c. adloff et al .
, ; zeus collab . , j. breitweg et al . , .
for a recent review , see e.g. r. yoshida , proceedings of dis2000 .
g. altarelli , n. cabbibo , l. maiani and r. petronzio , .
a diagram describing a dis process on a proton is shown in figure [ fig : kinematics ] . a lepton with mass @xmath43 and four - vector @xmath44 interacts with a proton with mass @xmath45 and four - vector @xmath46 through the exchange of a gauge vector boson , which can be @xmath47 , @xmath48 or @xmath49 , depending on the circumstances .
the four - vector of the exchanged boson is @xmath50 . with these notations
one can define the following variables , @xmath51 the meaning of the variables @xmath52 and @xmath35 is most easily realized in the rest frame of the proton . in that frame
@xmath52 is the energy of the exchanged boson , and @xmath35 is the fraction of the incoming lepton energy carried by the exchanged boson .
the variable @xmath53 is the squared center of mass energy of the gauge boson proton system , and thus also the squared invariant mass of the hadronic final state .
the variable @xmath54 is the squared center of mass energy of the lepton proton system .
the four momentum transfer squared at the lepton vertex can be approximated as follows ( for @xmath55 ) , @xmath56 the scattering angle @xmath57 of the outgoing lepton is defined with respect to the incoming lepton direction .
the variable which is mostly used in dis is the negative value of the four momentum transfer squared at the lepton vertex , @xmath58 one is now ready to define the other variable most frequently used in dis , namely the dimensionless scaling variable @xmath2 , @xmath59 to understand the physical meaning of this variable , one goes to a frame in which masses and transverse momenta can be neglected - the so - called infinite momentum frame . in this frame
the variable @xmath2 is the fraction of the proton momentum carried by the massless parton which absorbs the exchanged boson in the dis interaction .
this variable , defined by bjorken , is duly referred to as bjorken-@xmath2 .
the diagram in figure [ fig : kinematics ] describes both the processes in which the outgoing lepton is the same as the incoming one , which are called neutral current reactions ( nc ) , as well as those in which the nature of the lepton changes ( conserving however lepton number ) and which are called charged current processes ( cc ) . in the nc dis reaction
, the exchanged boson can be either a virtual photon @xmath60 , if @xmath1 is not very large and then the reaction is dominantly electromagnetic , or can be a @xmath48 which dominates the reaction at high enough @xmath1 values and the process is dominated by weak forces . in case of the cc dis reactions , only the weak forces are present and the exchange bosons are the @xmath49 .
the variables used to analyze diffractive scattering are introduced for @xmath0 dis .
a diagram for diffractive scattering in dis , where the diffracted state is separated from the scattered proton by a large rapidity gap ( lrg ) , is presented in figure [ fig : dis - diag ] and all the relevant four vectors are defined therein .
in addition to the usual dis variables , defined above , the variables used to describe the diffractive final state are , @xmath61 @xmath62 is the fractional proton momentum which participates in the interaction with @xmath60 .
it is sometimes denoted by @xmath63 .
@xmath30 is the equivalent of bjorken @xmath2 but relative to the exchanged object .
@xmath64 is the invariant mass of the hadronic final state recoiling against the leading proton , @xmath65 .
the approximate relations hold for small values of the four - momentum transfer squared @xmath66 and large @xmath24 , typical of high energy diffraction .
to describe diffractive dis , it is customary to choose the variables @xmath62 and @xmath66 in addition to the usual @xmath2 and @xmath1 in the cross section formula . the diffractive contribution to @xmath3
is denoted by @xmath67 and the corresponding differential contributions are @xmath68 the contribution from the longitudinal structure function is omitted for simplicity .
the structure function @xmath3 is related to the absorption cross section of a virtual photon by the proton , @xmath70 . for diffractive scattering , in the limit of high @xmath24 ( low @xmath2 ) , | the @xmath0 hera collider started operation in summer of 1992 .
this talk summarizes some of the highlights of physics results obtained since then and discusses their impact on our understanding of quantum chromodynamics ( qcd ) .
epsf 23.0 cm 17.0 cm -1.0 in -42pt plus 2 mm minus 2 mm # 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 _ eur .
phys .
j. _ * # 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 _ nucl .
phys .
proc .
suppl . _ * # 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
. phys
. jept _ * # 1 * ( # 2 ) # 3 # 1#2#3 _ sov .
phys .-
usp . _ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ zeit .
phys .
_ * # 1 * ( # 2 ) # 3 desy 00 - 116 + august 2000 + |
in global terms the financial correlations can be classified as correlations in space and correlations in time , though of course they are somewhat interrelated .
the first category so far studied involves for instance the correlations among the companies comprised by a single stock market @xcite , among a group of subjects @xcite and even between the different and geographically remote stock markets @xcite . from practical perspective
this type of correlations relates to the theory of optimal portfolios @xcite and risk management . in the context of the stock market
an important result of this study is that majority of eigenvalues in the spectrum of the correlation matrix agree very well with the universal predictions of random matrix theory @xcite .
locations of some of the eigenvalues differ however from these predictions and thus , similarly as in other physical systems @xcite , identify certain system - specific , non - random properties such as collectivity .
studying explicit correlations in time is at least as important because it is this type of correlations that directly reflects a nature of the financial dynamics . in general the character of those correlations
is however less understood and several related issues still remain puzzling .
it is for instance commonly accepted @xcite that the autocorrelation function of the financial time - series drops down to zero within few minutes which reflects a time - horizon of the market inefficiency . at the same time the correlations in volatility remain positive for many weeks . on short time - scales the return distributions
are definitely not levy stable @xcite but it turns out difficult to detect their convergence to a gaussian on longer time - scales as expected from the central limit theorem . in this connection it is also appropriate to mention a still poorly understood phenomenon of log - periodicity @xcite which seems to reflect existence of some very specific correlations on all time - scales . as a contribution towards resolving this sort of difficulties below we propose to use the concept of the correlation matrix such that it focuses entirely on the time - correlations and their potential existence
can parallelly be detected on various time - scales .
utility of such a procedure is illustrated on an example of high - frequency ( 15sec ) recordings of the deutsche aktienindex ( dax ) .
the entries of the correlation matrix are then constructed from the time - series of returns representing the consecutive trading days . as a result
such entries are labelled by the pairs of different days .
several striking observations are also made .
the relevant correlation matrix is defined as follows . to each element in a certain sequence @xmath0 of relatively long consecutive time - intervals of equal length @xmath1 labelled with @xmath2 one uniquely
assigns a time series @xmath3 , where @xmath4 @xmath5 is to be understood as discrete time counted from the beginning for each @xmath2 . in the present case @xmath2 is going to label the consecutive trading days and @xmath4 the trading time during the day .
similar methodology has already been successfully applied @xcite to extract the repeatable structures in the brain sensory response . if , as here , @xmath3 represents a price time - series than it is natural to define the returns @xmath6 time - series as @xmath7 where @xmath8 is the time lag .
the normalized returns , with the average value subtracted and its variance normalized to unity , are defined by @xmath9 where @xmath10 is volatility of @xmath11 and @xmath12 denotes averaging over time .
one thus obtains @xmath0 time series @xmath13 of length @xmath14 ( @xmath14=@xmath1 - 1 ) , _ i.e. _ an @xmath15 matrix @xmath16 .
then , the correlation matrix is defined as @xmath17 . by diagonalizing
@xmath18 @xmath19 one obtains the eigenvalues @xmath20 @xmath21 and the corresponding eigenvectors @xmath22 . in the limiting case of entirely random correlations the density of eigenvalues
@xmath23 defined as @xmath24 where @xmath25 is the number of eigenvalues of @xmath18 less than @xmath26 , is known analytically @xcite , and reads @xmath27 with @xmath28 , @xmath29 , and where @xmath30 is equal to the variance of the time series ( unity in our case ) .
our exploratory study along the above indicated line is based on the dax recordings with the frequency of 15 sec during the period between november 28th , 1997 and december 30th , 1999 . by taking the dax intraday variation between the trading time 9:03 and 17:10 which corresponds to @xmath31 one then obtains @xmath32 complete and equivalent time series representing different trading days during this calendar period .
( several days with incomplete recordings have been rejected . ) using this set of data we construct the @xmath33 matrix @xmath18 .
a characteristics of primary interest is the structure of eigenspectrum of @xmath18 .
the structure of eigenspectrum of a matrix is expected to be related @xcite to the distribution of its elements .
the distribution of such elements of @xmath18 corresponding to the above specified procedure and the data set is shown in the left panel of fig .
[ fig : fig1 ] clearly , this distribution is symmetric with respect to zero , a gaussian like ( dashed line ) on the level of small matrix elements , but sizably thicker than a gaussian on the level of large matrix elements .
as indicated in the figure , in the latter case , a power law with the index of about 5.6 ( consistent with the distribution of returns ) provides a reasonable representation .
this is far beyond the levy stable regime but points to the two ensembles of random matrices , the gaussian @xcite and the levy distributed @xcite , as interesting limits for formulating the null hypotheses . the resulting probability density of eigenvalues , shown in the right panel of fig .
[ fig : fig1 ] , displays a somewhat unexpected structure .
there exist two outliers significantly above @xmath34 ( for @xmath35 , @xmath36 ) which seem to indicate that the dynamics under consideration is not a pure noise , but instead that certain time specific repeatable structures in the intraday trading do take place .
however , it is even more astonishing that the bulk of the spectrum agrees very well with the bounds prescribed by purely random correlations .
this indicates that the neighbouring recordings in our time series of 15 sec dax returns share essentially no common information .
our previous experience @xcite teaches us that the probability density of eigenvalues of the correlation matrix is a sensitive measure of such effects . typically , a common information shared by neighbouring events results in an effective number ( @xmath37 ) of time points such that @xmath37 is significantly smaller than @xmath14 .
this immediately affects both @xmath34 and @xmath38 . in the present case
we see basically no such effect , particularly on @xmath39 side .
a whole nonrandomness can be here associated with the two largest eigenvalues . in order to identify the character of eigenstates associated with these eigenvalues we split the whole daily time interval considered into three equal subintervals . for each of them
( thus @xmath41 and @xmath42 ) .
the resulting distributions of matrix elements of @xmath18 and the corresponding probability densities of eigenvalues are shown in fig .
[ fig : fig2 ] .
it appears easy to see that separation of the two largest eigenvalues originates from the last subinterval .
consistently , it is also this subinterval which produces the thickest tails in the distribution of large matrix elements . an even more efficient way to visualize the differences among the eigenvectors is to look at the superposed time series of returns @xmath43 in this definition @xmath44 is used instead of @xmath45 for the reason of preserving normalization and the sign of @xmath45 to account for a possible coherence of the original signals . a collection of such superposed time series of returns for @xmath46 , 2 and 418 , and of the simple average of the original returns is shown in fig . [ fig : fig3 ] . for
a statistical value of @xmath47 , as it is illustrated by an example of @xmath48 ( ( c ) in fig [ fig : fig3 ] ) , @xmath49 basically does not differ from the simple average ( panel ( d ) in fig .
[ fig : fig3 ] ) .
the first two differ however significantly and indicate the existence of repeatable structures at the well defined instants of time through many days .
quite unexpectedly , the most collective signal , associated with the largest eigenvalue ( @xmath46 shown in ( a ) of fig . [
fig : fig3 ] ) , reflects the strongest synchronous dax activity precisely at around 14:30 ( probably in response to the north - american financial news release exactly at this time ) and not at the time when the wall street opens , nor even just before closing in frankfurt .
this last period of an enhanced activity also carries some signatures of synchrony but is entirely comprised by the second eigenstate .
= 12.0 cm = 8.0 cm = 8.0 cm another characteristics which carries the relevant information is the probability distribution of the eigenvector components @xmath45 . such a probability distribution evaluated from all @xmath47 s
is illustrated by the solid line in fig .
[ fig : fig4 ] .
it globally is well represented by a gaussian .
significantly different are the distributions of eigenvector components for our two outliers ; they are concentrated more at around zero but , at the same time , the tails of the distributions are thicker .
this effect is more pronounced for @xmath46 than for @xmath50 which indicates that fewer days ( @xmath2 s ) contribute to the strong signal seen at 14:30 for @xmath46 than just after 17:00 for @xmath50 .
= 12.0 cm this issue is quantified in more detail in fig .
[ fig : fig5 ] which shows how these two eigenvector components are distributed in magnitude over the trading days @xmath51 s ) incorporated in the present study . for @xmath46
such components assume large values indeed and even seem to be developing certain periodicity , which in the initial period considered here corresponds to about 20 trading days ( about one calendar month ) .
interestingly , it is this period which corresponds to our classic example of `` imprints of log - periodic self - similarity in the stock market '' @xcite .
perhaps the observed , equidistant in time , synchronous bursts of activity constitute one of the elements introducing a characteristic time - scale responsible for discrete scale invariance @xcite , an element which in natural way may contribute to the appearance of the log - periodic oscillations . finally , we find it appropriate to notice that the return probability density distributions associated with the above identified synchronous bursts of activity seem to be governed by a different law than what more global statistical analyses @xcite document . fig . [
fig : fig6 ] shows the tails of the return @xmath6 distribution ( triangles ) for the time period 14:2514:35 through all the dax 517 trading days considered here .
the statistics is of course poor but still it quite convincingly indicates much thicker tails as compared to the periods of normal activity , represented in this figure ( circles ) by the period 9:0314:25 . in the latter case
the pdf parameters are about consistent with those cited in the literature @xcite , while in the short time interval at around 14:30 we seem to be still in the lvy stable regime .
the present study sheds a somewhat new light on the issue of time - correlations in the dynamics of financial evolution .
it shows for instance that the consecutive returns carry essentially no common information even when probed with the frequency of 15 sec . such a conclusion sounds somewhat in contrast to a common belief , based on autocorrelation function analysis , that the market inefficiency time - horizon extends over few minutes .
there is however perhaps no contradiction between our conclusion and the time - lag dependence of the autocorrelation function .
this function remains positive for about few minutes indeed , but already after 15 sec it drops down by at least one order of magnitude . on the other hand , however , at the well defined short periods of time during the intraday trading , there exist clearly identifiable synchronous repeatable bursts of activity ( for dax at 14:30 ) whose related return probability density functional develops significantly larger values on the level of rare events than during periods of the normal activity .
this , together with the fact that such events are strongly correlated in time , may constitute a principal reason for an observed extremely slow convergence to a gaussian of the global return distribution on long - time scales .
it would now be very interesting to perform similar study for the other stock market indices as well . | a novel application of the correlation matrix formalism to study dynamics of the financial evolution is presented .
this formalism allows to quantify the memory effects as well as some potential repeatable intradaily structures in the financial time - series .
the present study is based on the high - frequency deutsche aktienindex ( dax ) data over the time - period between november 1997 and december 1999 and demonstrates a power of the method . in this way
two significant new aspects of the dax evolution are identified : ( i ) the memory effects turn out to be sizably shorter than what the standard autocorrelation function analysis seems to indicate and ( ii ) there exist short term repeatable structures in fluctuations that are governed by a distinct dynamics .
the former of these results may provide an argument in favour of the market efficiency while the later one may indicate origin of the difficulty in reaching a gaussian limit , expected from the central limit theorem , in the distribution of returns on longer time - horizons . epsf , , , , |
brightest cluster galaxies ( bcgs ) are luminous ellipticals found near the centres of galaxy clusters and groups .
they are further subdivided into giant ellipticals ( ge ) , d galaxies which have shallower surface brightness profiles than ges , and finally cd galaxies which have extended low surface brightness envelopes ( mathews , morgan & schmidt 1964 ; schombert 1987 , 1988 ) .
bcgs are often associated with powerful radio sources ( mathews _ et al .
_ 1964 ) , and extended cluster x - ray emission ( jones _ et al .
_ 1979 ) . in regular clusters ,
bcgs are invariably at the centre of the cluster potential , as defined by the cluster x - ray emission , galaxy distributions , and strong / weak gravitational lensing ( e.g. miralda - escude 1995 ) , and their velocities are usually close to the mean cluster velocity .
bcgs are often observed to have extremely high globular cluster specific frequencies ( e.g. harris , pritchet , & mcclure 1995 ; bridges _ et al .
_ 1996 ) .
the origin of bcg halos is one of the most puzzling aspects of galaxy formation , and it is unclear whether they are formed at the same time as the galaxy proper , as proposed by merritt ( 1984 ) , or are accreted later .
if accreted later they might form via dissipational processes , such as star formation in cooling flows ( fabian & nulsen 1977 ) , or by purely stellar dynamical processes , such as galactic cannibalism ( hausman & ostriker 1978 ; mcglynn & ostriker 1980 ) , accretion of tidally stripped material ( richstone 1976 ) or `` galaxy harassment '' ( moore _ et al .
_ 1996 ) . for a review of bcg formation ,
see tremaine ( 1990 ) ; see also garijo , athanassoula , & garcia - gomez ( 1997 ) , and dubinski ( 1998 ) .
the kinematics and metallicities of bcg halos provide important clues as to how they form .
if they form in a dissipational process , then we might expect a low velocity dispersion , as the energy of the material from which the stars form has been radiated away .
if bcgs form in a dissipative collapse , we would expect to see metallicity gradients in the halo ( though metallicity gradients are not unique to dissipative formation models ) .
if the halos form out of material from other cluster galaxies via dissipationless processes , then we might expect the halos to reflect the kinematics of the cluster as a whole , with high velocity dispersion and little net rotation ; metallicity gradients would be shallower if bcgs experience significantly more mergers than other cluster galaxies , since mergers are expected to weaken metallicity gradients .
if the halos are primordial , and formed in a process involving violent relaxation , then the kinematics will reflect the underlying mass distribution .
while there is considerable information on the central kinematics of ellipticals , there have been few studies of the halo kinematics at very large radii from optical spectroscopy .
most such studies have concentrated on normal ellipticals , and have found that the velocity dispersion profiles are either flat or decreasing with radius ( e.g. saglia _ et al .
_ 1993 ; carollo _ et al .
_ 1995 ; statler , smecker - hane , & cecil 1996 ) .
similar results have been found in the even fewer studies of bcgs ( e.g. carter _ et al .
_ 1981 , 1985 ; tonry 1984 , 1985 ; heckman _ et al .
_ 1985 ; fisher , illingworth , & franx 1995 ) . to date ,
one cd ( ic 1101 in abell 2029 : dressler 1979 ; fisher _ et al .
_ 1995 ; sembach & tonry 1996 ) , and two dumbell galaxies ( abell 3266 : carter _ et al .
_ 1985 , and ic 2082 : carter _ et al .
_ 1981 ) , have been found to have _ rising _ dispersion profiles .
an alternative approach to the study of the kinematics of the outer halos of cd and other galaxies is the use of globular clusters and planetary nebulae as tracers . because of the geometric constraints of current multi - slit and fibre fed spectrographs
, this technique can be used only in the outer halos , and because of magnitude limits can not be applied at distances beyond two or three times the distance of the virgo cluster . nevertheless for m87 ( mould _ et al .
_ 1990 ; cohen & ryzhov 1997 ) and ngc 1399 ( grillmair _ et al .
_ 1994 ; kissler - patig _ et al .
_ 1998 ; minniti _ et al .
_ 1998 ) , globular cluster velocities have shown conclusively that the m / l ratio increases outwards in the halo , and thus that there are substantial dark matter halos in both galaxies .
( 1994 ) confirm this result for ngc 1399 for a sample of planetary nebulae .
interestingly , cohen & ryzhov ( 1997 ) find that the globular cluster velocity dispersion in m87 rises outwards , and in ngc 1399 the velocity dispersions of the planetary nebulae and globular clusters match smoothly onto the velocity dispersion of the fornax cluster itself .
these bcgs at least seem to merge smoothly into their host clusters , and the bcg and cluster dynamics seem closely linked ( see discussion in freeman 1997 ) .
the evidence from the distribution of lensed arcs in more distant clusters points to the dark matter being more centrally concentrated than the x - ray gas ( miralda - escude & babul 1995 ) . in this case
the dark matter will affect the dynamics of the cd halo significantly .
we have embarked on a program to obtain very deep optical spectroscopy of bcgs . here
we present long - slit spectra for ngc 6166 , ngc 6173 , and ngc 6086 , obtained at the 2.5 m int in la palma in may 1996 .
the properties of these galaxies and their host clusters are listed in table [ tab1 ] .
ngc 6166 is a luminous cd and a classic multiple - nucleus galaxy ( though the ` nuclei ' are probably not bound to ngc 6166 ) , centrally located in a rich cluster with a large cooling flow ( @xmath9 160 m@xmath10/yr : allen & fabian 1997 ) , and a regular , symmetrical x - ray appearance ( buote & canizares 1996 ) .
abell 2197 is 1.3 degrees north of abell 2199 ( gregory & thompson 1984 ) , and is a sparser cluster with an irregular x - ray morphology dominated by two main concentrations ( muriel , bohringer , & voges 1996 ) .
ngc 6173 is a d galaxy found at the centre of one of these concentrations ( subclusters ) , and appears to have a significant peculiar velocity ( i.e. relative to the cluster mean velocity ) .
abell 2162 is probably best described as a poor cluster or compact group ; we have not been able to find any x - ray data for this cluster .
however , ngc 6086 is classified as a cd galaxy based on its surface brightness profile , and does not have a peculiar velocity .
thus , our three galaxies / clusters span a wide range of properties .
[ tab1 ] cccccccccc galaxy / cluster & galaxy & m@xmath11 & @xmath2 & v@xmath12 & v@xmath13 & @xmath14 & bm & r & x - ray + & type & & & ( km s@xmath0 ) & ( km s@xmath0 ) & & & + + & & & our & sample & & & & + + n6166/a2199 & cd & @xmath1522.7 & 326 & 9293 & 9063 & 823 & i & 2 &
cf , reg + n6173/a2197 & ge & @xmath1522.4 & 261 & 8800 & 9134 & 550 & iii & 1 & irr + n6086/a2162 & cd & @xmath1521.8 & 304 & 9547 & 9795 & 302 & ii - iii & 0 & ? + + + ic 1101/a2029 & d & @xmath1523.5 & 359 & 23399 & 23169 & 1436 & i & 2 & cf , reg + sersic 40 - 6/a3266 & cd & & 327 & 17914 & 17802 & 1186 & i -
ii & 2 & cf , reg + ic2082/as0463 & d & & 265&12051&12035&844 & i - ii & 0 & ?
+ [ tab1 ] ngc 6166 has previously been observed spectroscopically by tonry ( 1985 ) and fisher _ et al . _
( 1995 ) with 2 and 3 position angles respectively ( each include the major axis ) .
unfortunately , both studies only extend to 10@xmath1520 arcsec from the galaxy centre , and over this limited range both find a flat or slightly decreasing velocity dispersion profile ( we show below that this is consistent with our data , which extend much further out ) .
fisher _ et al .
_ find that ngc 6166 does not have significant rotation along either the major or minor axes out to 20 arcsec radius .
we have only been able to find central velocity dispersions and line strengths for the other two galaxies ( e.g. burstein _ et al .
_ 1987 ; mcelroy 1995 ) .
table [ tab2 ] contains details of the observations .
spectra are in each case taken with the slit aligned with the major axis to maximise the signal to noise ratio .
the spectra include the mg and fe absorption lines , as well as [ oiii ] and [ ni ] emission lines ( for ngc 6166 only ) .
we also obtained template spectra of 28 stars , with spectral type ranging from g6 to m1 , and luminosity class iii , iv and v. figure [ fig : censpec ] shows the central spectra for each of the three galaxies ; for each spectrum , the best - fitting stellar template is also shown ( see below ) , as well as the bandpass definitions for 2 g , 52 , and 53e .
preliminary data reduction was done using standard figaro routines , and followed the procedures of carter , thomson and hau ( 1998 ) .
this involves debiassing , flatfielding and wavelength calibrating the data .
a slope in the spectrum due to detector rotation was corrected using the s - distortion correction routines in figaro .
the template spectra were re - binned to zero velocity , using their published radial velocities , mainly from wilson ( 1953 ) .
[ tab2 ] cc + telescope & isaac newton telescope ( 2.5 m ) + instrument & intermediate dispersion spectrograph + observing dates & 1016/06/1996 + camera & 235 mm + grating & r1200y + dispersion & 35 / mm + detector & tektronix 1024 ccd + resolution & 2 + spectral coverage & 49305730 + slit width & 3.0 arcsec + average seeing & 1.3 arcsec + slit pa & 41/140/0 deg@xmath16 + exposure time & 18.75/6.25/6.33 hrs@xmath16 + @xmath16 for ngc 6166/6173/6086 respectively .
[ fig1 ] following van der marel & franx ( 1993 ) , the line - of - sight velocity distributions ( losvds ) are modelled as gaussians with small hermite deviations , parameterised as mean velocity @xmath17 , dispersion @xmath18 , and hermite moments @xmath19 of order @xmath20 .
the parameters are recovered by the program kinematics written by h - w .
rix ; for more details please see van der marel & franx ( 1993 ) , rix & white ( 1992 ) , and hau , carter , & balcells ( 1998 ) .
the gauss - hermite moments 3 and 4 are closely related to the skewness and kurtosis of the losvd respectively , and are therefore easy to visualise .
large 3 with opposite sign to the mean velocity is typical of rapidly - rotating systems , whilst for spherical two - integral models , 4 is proportional to the velocity dispersion anisotropy @xmath21 ( van der marel & franx 1993 ) .
the higher moments are more difficult to interpret .
however , they are useful for identifying structures in the residuals of the model fit to the data
. we measure and plot gauss - hermite parameters up to 6 .
the observed wavelength range is well known to give template mismatch problems ( van der marel & franx 1993 ) . in order to minimize template mismatch ,
an optimal stellar template is employed for the direct fitting .
this is estimated as the best linear combination of the 28 template spectra which when broadened with a gaussian profile minimizes @xmath22 , the difference between the combination and the galaxy spectrum , after an initial guess of the velocity for each template and the dispersion , ( rix & white 1992 ) .
this optimal template is then used for recovering the losvd for the galaxy spectrum , modelled up to 6 .
this procedure is done for each of the individual spectra in each galaxy , thus reducing radial template mismatch due to abundance gradients .
the galaxy nucleus is located by a gaussian fit to the light profile of the central few pixels .
monte carlo experiments show that 2500 counts / pixel or higher gives satisfactory results for the direct fitting ( see appendix a ) .
our procedure allows the determination of an individual optimal composite template for each data point , to ensure that this procedure does not introduce systematic effects into our profiles we repeated the analysis for ngc 6166 using the composite template for the central spectrum for all data points .
there are no systematic differences in zeropoint or overall trend in any of the kinematic values derived .
subjectively , we estimate that the scatter is higher , justifying our decision to use individually determined optimal templates
. emission lines could affect the measurement of absorption - line indices if they fall within one of the bandpasses , and also can potentially introduce systematics in the absorption line profile analysis .
significant [ oiii ] and [ ni ] emission are observed in the nuclear region @xmath23 of ngc 6166 , but not in ngc 6173 or ngc 6086 .
thus , for ngc 6166 , pixels within a wavelength difference corresponding to 3@xmath18 from an emission line are excluded from the line - profile fitting .
these regions are : @xmath24@xmath25 ( [ oiii ] ) , @xmath26@xmath27 ( [ ni ] ) , and @xmath28@xmath29 ( strong sky line ) .
experimentation shows that excluding a few pixels has a negligible effect on the recovery of the losvd as kinematics are measured over the entire spectrum .
for spectra more than 11 arcsec from the nucleus where no emission is present , only the region affected by the strong sky line is excluded , allowing fitting over the entire spectrum .
another difficulty with ngc 6166 is the presence of companion nucleus b , which is at a pa of 62with respect to the central cd , and contributes significantly to the light in our slit between @xmath9 5@xmath1520 arcsec from the galaxy centre along the ne side .
kinematics was not designed to recover two losvds , so we restrict our analysis to the sw side of the galaxy .
the parameters recovered by the gauss - hermite analysis are presented in figures [ fig : hermplot](a ) , ( b ) , and ( c ) , for ngc 6166 , ngc 6173 , and ngc 6086 respectively .
_ central velocity dispersions
_ we can compare our central velocity dispersions , @xmath2 = 320 @xmath30 5 , 295 @xmath30 5 , and 325 @xmath30 8 km s@xmath0 for ngc 6166 , 6173 , and 6086 respectively , with previous determinations .
ngc 6166 has been the best - studied of the three : malumuth & kirshner ( 1981 ) , tonry ( 1985 ) , burstein _ et al . _
( 1987 ) , and fisher _ et al . _
( 1995 ) , find @xmath2= 350 @xmath30 35 , 303 , 326 , and 295 @xmath30 6 km s@xmath0 respectively . for ngc 6173 , burstein _ et al .
( 1987 ) and mcelroy ( 1995 ) find @xmath2= 261 and 240 km s@xmath0 respectively , both values lower than ours .
for ngc 6086 , burstein _ et al . _
( 1987 ) and mcelroy ( 1995 ) find @xmath2= 304 and 325 km s@xmath0 respectively .
overall , the agreement is good , and the present data are of much higher quality than any previous data . _ velocity dispersion profiles _ from figures [ fig : hermplot](b ) and ( c )
, we see that both ngc 6173 and ngc 6086 have velocity dispersion profiles that decline outwards .
such behavior is found in most normal ellipticals .
in contrast , figure [ fig : hermplot](a ) shows that the velocity dispersion profile of ngc 6166 _ increases _ from @xmath9 325 km s@xmath0 at the centre to @xmath9 450 km s@xmath0 at 35 arcsec along the major axis .
tonry ( 1985 ) and fisher _ et al . _
( 1995 ) both have limited dispersion profiles for ngc 6166 , and find a flat and slightly decreasing dispersion profile respectively , out to @xmath9 20 arcsec along the major axis .
this is consistent with what we find , since our dispersion only starts rising significantly beyond @xmath9 20 arcsec . in appendix
a we analyse possible systematic errors introduced into the kinematic parameters by noise and imperfect sky subtraction .
we find that there is a bias towards low velocity dispersion when the signal - to - noise is poor , but this does not significantly affect the points plotted in figure 2(a - c ) . from figure
[ fig : hermplot](a ) , ngc 6166 shows modest major - axis rotation , and a weighted least - squares fit yields a slope of 1.07 @xmath30 0.12 km s@xmath0 arcsec@xmath0 , or a rotation of @xmath9 45 km s@xmath0 at 40 arcsec .
tonry ( 1985 ) found a rotation amounting to 4 km s@xmath0 arcsec@xmath0 at pas of 62and 102 , while fisher et al ( 1995 ) find no significant rotation at pas of 31 , 35and 125 .
ngc 6173 and ngc 6086 both show evidence of rotation in the core , decoupled from the kinematic properties further out . in ngc 6173 ( figure [ fig : hermplot](b ) ) , a linear least - squares fit gives a velocity gradient of @xmath152.56 @xmath30 0.27 km s@xmath0 arcsec@xmath0 for @xmath31 arcsec , with no significant rotation outside this , and in ngc 6086 ( figure [ fig : hermplot](c ) ) , similar fits give a velocity gradient of 0.86 @xmath30 0.48 km s@xmath0 arcsec@xmath0 for @xmath32 arcsec and @xmath151.33 @xmath30 0.13 km s@xmath0 arcsec@xmath0 for @xmath33 arcsec .
although the core rotation is only marginally significant , it does appear to be counter - rotating .
ngc 6173 is a shell galaxy , and both the shells and higher core rotation could be a result of a merger or interaction .
however , in both galaxies , there is no significant asymmetry in the line profiles ( i.e. changes in @xmath34 ) associated with the peculiar kinematics , as expected from classical kdcs like ic 1459 or ngc 2865 ( hau , carter , & balcells 1998 ) .
also , statler ( 1991 ) proposes that kdcs might also arise from streaming in a triaxial potential , without a merger / interaction . the parameter @xmath1 measures symmetric deviations from a gaussian profile and is related to the velocity anisotropy .
@xmath1 is significantly positive , with no radial dependence , in all of our galaxies . although it is not straightforward to translate @xmath1 to the velocity anisotropy parameter @xmath21 ( @xmath35 ) , positive values , at least outside the nucleus ,
generally indicate a bias towards radial orbits ( gerhard 1993 ; gerhard _ et al .
_ 1998 , rix _
the most important result from the line - profile analysis is that there is no change in @xmath1 associated with the rising dispersion in ngc 6166 , indicating that the increase in velocity dispersion is not associated with a change in velocity anisotropy towards tangential orbits .
@xmath1 has a magitude of 0.10 , 0.05 & 0.05 in ngc 6166 , 6173 & 6086 respectively .
template mismatch can introduce systematics in @xmath1 by either increasing or decreasing its magnitude ; see , for example , van der marel _
( 1994 ) , and hau , carter , & balcells ( 1998 ) .
the @xmath22 fits in ngc 6173 & ngc 6086 appear to be near perfect .
inspection of the residuals for ngc 6166 shows that the fits are less perfect than for the other two galaxies , but are satisfactory .
we feel that template mismatch is not a major problem for all three galaxies as the optimal template is estimated from 28 template stars . in order to estimate the effect of template mismatch on the absolute value of @xmath1 ( and other kinematic parameters ) , in appendix a
we derive the kinematic parameters for ngc 6166 for a range of individual templates .
none of these templates fit the galaxy data as well as the optimal template . given an extreme range of templates , we find zero point differences of up to 0.1 in @xmath1 , @xmath36 and @xmath37 ; 0.2 in @xmath34 , and 50 km / s in velocity dispersion .
however these differences are largely in the zero points for stars at the extreme end of the range of spectral types of the observed templates ( g6 iii and m1 iii ) .
in particular low @xmath1 is found only when an m1 iii template is used , this template does not fit the data well and contributes little to the optimal composite templates . although template mismatch may in principle shift the zero - point in @xmath1 , it can not explain the significantly positive @xmath1 seen in our data . in appendix
a we also analyse the effect of noise on the recovery of @xmath1 when the magnitude of the real value of @xmath1 is large .
we find that there are no systematic biases unless @xmath38 , when @xmath39 is systematically underestimated at low signal to noise .
thus the only possible bias in our results would be to underestimate @xmath1 in the outer regions of ngc 6166 , although there is no evidence that this has happened , we can not absolutely rule it out .
van der marel & franx ( 1993 ) give formulae which relate @xmath1 to @xmath21 , but they are model - dependent .
to better quantify the anisotropy parameter @xmath21 , we need to solve for the @xmath40 ratio using all of the available kinematic and photometric data .
such techniques have been developed by rix _
( 1997 ) , using schwarzchild s method of populating orbits , and by dehnen ( 1995 ) for the axisymmetric case in order to model m32 . in summary ,
the rising velocity dispersion profile in ngc 6166 is not due to a change in velocity anisotropy , but rather , reflects an increasing @xmath40 ratio .
this indicates that ngc 6166 possesses a massive dark halo .
significant [ oiii ] and [ ni ] emission is observed in the centre of ngc 6166 . in this section
we present the line - of - sight velocity distribution of the ionized gas inferred by the profile of the [ oiii ] and [ ni ] emission lines , obtained directly from the residual map of the gauss - hermite analysis , described in the previous section .
this method is very successful and is more objective than fitting the baseline by hand .
the results are plotted in figure [ fig : gaskin ] , with the stellar rotation curve from figure [ fig : hermplot](a ) over - plotted .
we see that the [ oiii ] and [ ni ] emission is very concentrated to the galaxy centre , with spatial fwhm of 1.9 and 3.0 arcsec respectively .
the spatial fwhm of template stars is about 2.3 arcsec , thus the [ oiii ] emission region is unresolved but that of the [ ni ] may be marginally resolved . at the galaxy centre ,
the [ oiii ] and [ ni ] have velocity dispersions @xmath18 of @xmath41 and @xmath42 km s@xmath0 respectively , which is very similar to the central stellar velocity dispersion ( figure 2a ) .
in this section the absorption line indices are presented for the three galaxies .
mg@xmath7 , 52 , and 53e are extracted from individual , de - redshifted spectra following the definitions and recipes of faber ( 1985 ) , and adopting the passband definitions of worthey ( 1994 ) ; see appendix b for more details .
we first discuss the local mg@xmath43fe@xmath44 relations for each galaxy , then present the mg@xmath7 radial profiles .
we discuss the significance of these results in our conclusions . as discussed in appendix b , there is some uncertainty regarding the zero - point of our measured indices in the lick / ids scale , due to the fact that only one star with lick indices ( hr 5227 ) has been observed , and also that the grating settings for the galaxy and the template star observations were different .
however recent observations by trager _
( 1998 ) for the nucleii of ngc 6166 and ngc 6086 do agree within their admittedly rather large error bars .
we also find that our 2 g measurements agree well with those measured by cardiel _
( 1997 ) , if both are not zero - point corrected ( figure [ fig : uscardiel ] ) . as most of our discussion
is not based on the absolute values of the indices , throughout this section we present line indices without zero - point corrections . in figure
[ fig : fevsmg2 ] , we plot @xmath45fe@xmath44 [ @xmath46 ( 52 + 53e)/2 ) ] against 2 g for the three galaxies .
all three galaxies appear to have [ mg / fe ] ratios higher than the model values with solar [ mg / fe ] abundance , but there are uncertainties in our zero - points .
notice further from figure [ fig : fevsmg2 ] that ngc 6086 and ngc 6166 are more ` enhanced ' than ngc 6173 .
this is interesting , given that the first two galaxies are centrally - located cds , while ngc 6173 is best described as a ge in a subcluster in abell 2197 . similarly , m87 is offset from the non - bcgs in the data of davies , sadler , & peletier .
hau ( 1998 ) has pointed out that in galaxies of high velocity dispersion , a positive value of @xmath1 , such as might result from radial anisotropy , leads to more absorption being lost from the passband that the fe indices are measured from , and thus to an artificially high value of [ mg / fe ] .
this systematic error might account for as much as a third of the offset between ngc 6166 and ngc 6173 .
however we still find that our galaxies with cd morphology have higher [ mg / fe ] compared with ngc 6173 , which is the only normal bright elliptical in our sample . if further observations confirm that ngc 6173 is typical ( despite uncertainty in the relative zero - points , the data of davies _ et al . _
( 1993 ) support this ) , then there is a difference between cds and normal bright ellipticals .
bender ( 1996 ) uses the offset of [ mg / fe ] to higher values in ellipticals when compared with spirals as an argument against the idea that ellipticals form from the mergers of objects similar to _ present - day _ spirals ; any such argument is clearly stronger for cds .
bender ( 1996 ) and davies ( 1996 ) feel that the most plausible explanation for the non - solar [ mg / fe ] ratios in ellipticals is a short burst of star formation , @xmath45 1@xmath152 gyr .
this would produce a population enriched in mg from rapid type ii supernovae from massive stars , while few stars would be formed from fe - enriched gas formed in longer - lived type ia supernovae .
see section 6 for further discussion on these topics .
we first compare our mg@xmath7 profile for ngc 6166 with that of cardiel _ et al . _
( 1997 ) in figure [ fig : uscardiel ] .
we have only plotted the sw side of the major axis because of contamination from the secondary nucleus on the ne side .
neither set of data has had any zeropoint corrections applied ( see appendix b ) . out to @xmath9 15 arcsec
, the agreement is very good . beyond 15 arcsec ,
our 2 g is significantly higher than that of cardiel _ et al . _
the source of this discrepancy is unclear , however our signal - to - noise is higher than that of cardiel _ et al .
_ their three outermost points have 2 g values @xmath45 0.2 , lower than any of the ellipticals studied by davies _
these extreme values seem unlikely , and we are confident of the reliability of our 2 g profiles .
the sensitivity of our line strength measurements to sky subtraction errors needs to be considered .
the sky spectrum that we subtract is averaged over many spatial pixels and is relatively low noise .
the only feature in the sky spectrum are the night sky emission line at 5577 and ( much weaker ) 5199 , apart from this the sky is well represented by a constant continuum . at these redshifts
the [ ni ] line at 5199 does not fall in the 2 g band or its continuum bands , and so can be neglected .
therefore the sky in the region of all of the line and continuum bandpasses can be represented by a constant , and the error in this constant can be determined from the maximum residuals seen in the [ oi ] line at 5577 in the sky subtracted spectra , and the ratio of the peak intensity in this line to the continuum in the sky spectra .
inspection of the sky line residuals in the outer regions of our profile show that sky subtraction errors do not exceed 1% of the signal level , thus systematic errors in our 2 g values will not exceed 0.01 mag . in figure
[ fig : mg2prof ] , we present the major axis profiles of 2 g , for ngc 6166 , 6173 , and 6086 . outside
35 arcsec , the 2 g profiles are well - fit by power - laws of slope ( d(2g)/d(log @xmath47 ) ) @xmath150.080 , @xmath150.059 , and @xmath150.073 for ngc 6166 , 6086 and 6173 respectively ( inside 3 arcsec , seeing effects are likely to be important ; the flattening of 2 g in ngc 6166 within a larger radius of @xmath9 5 arcsec may be due to star formation inside the emission - line region in this galaxy ) .
these slopes are fairly similar to each other , and are also similar to the average 2 g slope of @xmath150.058 found by cardiel _
( 1997 ) for eight bcgs without cooling flows or optical emission lines .
gorgas , efstathiou , & aragon - salamanca ( 1990 ) also measured shallower gradients for two bcgs ( @xmath150.051 and @xmath150.021 for the cds in 0559 - 40 and pks 2354 - 35 ) .
however , examination of their figure 9 shows that these gradients are based on only three points , and the outermost point in each case has large uncertainty .
cardiel _ et al . _
( 1997 ) find that the 2 g slopes are shallower in bcgs with optical line - emission , inside the emission - line region ; our data for ngc 6166 are consistent with this .
_ attribute this flattening to star formation from cooling flow gas near the galaxy centre .
this makes comparison of bcg 2 g data difficult , unless one is careful to compare slopes outside emission - line regions in all cases .
we can also compare our 2 g slopes with those of non - bcg ellipticals .
couture & hardy ( 1988 ) found a mean 2 g gradient of @xmath150.053 @xmath30 0.015 for six early - type galaxies ; davies , sadler , & peletier ( 1993 ) find a mean slope of @xmath150.059 @xmath30 0.022 for 13 normal ellipticals ; gorgas _ et al . _
( 1990 ) obtain a mean slope of @xmath150.058 @xmath30 0.027 for 16 ellipticals ; davidge ( 1992 ) found a steeper gradient of @xmath150.081 @xmath30 0.01 for 11 ellipticals .
gonzalez & gorgas ( 1998 ; see also gonzalez & gorgas 1997 ) find a mean 2 g gradient of @xmath150.055 @xmath30 0.025 for 109 early - type galaxies .
thus , our three bcgs have slopes consistent with those found in non - bcg ellipticals .
it seems likely that these bcgs have not experienced significantly more mergers than normal ellipticals , as mergers are expected to dilute abundance gradients ( white 1980 ) .
_ our main results are : _ * ( 1 ) : * we find small rotation along the major axes of ngc 6166 , ngc 6173 , and n6086 , with v@xmath48 @xmath49 50 km s@xmath0 at the outermost observed major axis radii ( 3040 arcsec ) in the three galaxies ; the corresponding v/@xmath18 ranges from 0.020.18 .
this small rotation is consistent with the nearly complete lack of rotation found near the centres of a sample of 13 bcgs by fisher _
this is consistent with merger models of bcg formation , but data at larger radii are urgently needed , since numerical simulations predict that mergers should contain significant amounts of angular momentum beyond 2@xmath153 r@xmath50 ( e.g. hernquist 1993 ) .
ngc 6173 and ngc 6086 have larger velocity gradients near the galaxy centres , and ngc 6086 appears to have a counter - rotating core .
these cores may arise from a previous merger / interaction ( supported by the shells seen in ngc 6173 ) , or may be due to streaming in a triaxial potential ; further observations at other position angles will be needed to decide between these two alternatives ( statler 1991 ) . * ( 2 ) : * ngc 6086 and ngc 6173 have slowly declining velocity dispersion profiles .
in contrast , ngc 6166 has a rising velocity dispersion : @xmath18 increases from @xmath9 325 km s@xmath0 at the centre to @xmath9 440 km s@xmath0 at 35 . *
( 3 ) : * the gauss - hermite 4 moment is positive and roughly constant at 0.05 at all radii along the major axis of ngc 6173 and ngc 6086 , and 0.1 along the major axis of ngc 6166 .
template mismatch is unlikely to account for all of this offset , and we conclude that there is possibly radial anisotropy present
. we will confirm and quantify this suggestion with detailed modelling in a later paper . *
( 4 ) : * we find [ oiii ] and [ ni ] emission concentrated within 5 arcseconds ( 3h@xmath0 kpc ) of the centre of ngc 6166 , with a central velocity dispersion of @xmath9 350 km s@xmath0 , similar to the central stellar velocity dispersion .
the extent of the [ oiii ] emission region is unresolved , but that of the [ ni ] may be marginally resolved . *
( 5 ) : * we find that [ mg / fe ] is larger in ngc 6166 and ngc 6086 than in ngc 6173 .
although positive @xmath1 can lead to systematic errors in [ mg / fe ] , the magnitude of the offset appears too large to be explained by such an error .
the difference in [ mg / fe ] between galaxies with cd morphology and normal bright ellipticals , if confirmed by further observations , would present a difficulty for models in which the cds were formed from mergers of normal galaxies of any type . *
( 6 ) : * the 2 g gradients outside 35 arcsec range from @xmath150.06 to @xmath150.08 for the three galaxies .
these slopes are consistent with those found for bcgs without cooling flows or emission lines by cardiel _
( 1997 ) , and for non - bcg ellipticals .
more data are needed to definitively compare bcgs with non - bcg ellipticals in this regard .
we defer to a forthcoming paper a detailed dynamical modelling of these three galaxies , but we note that ( 2 ) and ( 3 ) together imply that the m / l ratio increases outwards in the ngc 6166 halo , and likely in the other two galaxies as well . further support for a dm halo in ngc
6166 comes from x - ray data , e.g. buote & canizares ( 1996 ) .
do these results favour one bcg formation model over others ?
there are five main scenarios for bcg formation : * ( i ) * deposition of cooling flow gas from the icm ( e.g. fabian 1994 ) ; * ( ii ) * cannibalism of other cluster galaxies ( ostriker & tremaine 1975 ) ; * ( iii ) * tidal stripping of other cluster galaxies , either by the cluster tidal field or by the bcg itself ( richstone 1976 ; moore _ et al .
_ 1996 ) ; * ( iv ) * earlier origin in smaller groups / subclusters , which later merged to form the present - day cluster ( merritt 1984 , 1985 ; tremaine 1990 ) ; and * ( v ) * dissipative collapse models ( larson 1975 ; carlberg 1984a , b ) .
these scenarios are not necessarily mutually exclusive ; for instance , many authors have suggested that mergers / cannibalism create the central body of the cd , while tidal stripping accounts for their extended halos .
dubinski ( 1998 ) has run an impressive simulation of cd formation , using a large cosmological n - body simulation of cluster collapse with hierarchical merging .
his ` cluster ' has a mass of 1.0 @xmath51 10@xmath52 m@xmath10 and a line - of - sight velocity dispersion of 550 km s@xmath0 , and he notes that it would be classified as a poor cluster or large group by observers .
he finds that the final central giant elliptical has a small rotation of @xmath9 50 km s@xmath0 , is radially anisotropic in the outer regions , is aligned in position angle with the ` cluster ' galaxy distribution , and has a slowly declining major - axis velocity dispersion profile .
these properties agree with those observed in many bcgs , including our sample described in this paper .
all three of our galaxies have little or no rotation , and have significantly positive @xmath1 values which generally indicate radial anisotropy .
ngc 6086 and ngc 6173 have velocity dispersion profiles similar to dubinski s remnant , and ngc 6173 and ngc 6166 are aligned ( to within 20 degrees ) of their host cluster / subcluster galaxy distributions and x - ray isophotes ( carter & metcalfe 1980 ; dixon , godwin , & peach 1989 ; muriel , bohringer , & voges 1996 ; buote & canizares 1996 ) .
however , more such simulations are needed to see if , among other things , remnants with increasing dispersion profiles and extended halos can be created .
garijo , athanassoula , & garcia - gomez ( 1997 ) have run a larger set of simulations of cd formation , with a variety of initial conditions .
some of the simulations produce objects with extended halos and radial anisotropy , and all of the final galaxies have declining velocity dispersion profiles .
note , however , that these simulations were run with all of the mass initially bound to galaxies , probably not physically realistic .
the radial anisotropy found by both authors is a result of the ` cosmological ' mergers that are being modelled , where the initial filamentary structure means that merging galaxies enter the central object on mainly radial orbits .
further constraints can be placed on formation models if one considers the stellar abundance gradients .
our three galaxies all show modest gradients , comparable with normal bright ellipticals ( e.g. davies , sadler , & peletier 1993 ) .
these gradients are not as steep as predicted by simple dissipational collapse models ( e.g. larson 1975 ) , however they are steeper than we would expect from hierarchical dissipationless merger models , which predict no gradients .
it is not clear what the net effect of gaseous mergers will be , since on the one hand mergers are expected to dilute existing abundance gradients ( white 1980 ) , while on the other hand central star formation induced by the merger could augment any gradient ( see for instance mihos & hernquist 1994 ) . in any case
the gradients point to a formation history not too different from that of normal bright ellipticals . the [ mg / fe ] ratio in ngc 6166 and ngc 6086 , and the difference between these two bcgs and ngc 6173 , provide important new clues as to the differences between cds and normal bright ellipticals .
if this difference is confirmed by further observations , this is a second property , alongside the globular cluster specific frequency , in which ellipticals differ from spirals and in which cds are even more extreme .
fisher _ et al . _
( 1995 ) and bender ( 1996 ) find that [ mg / fe ] is correlated with velocity dispersion , so perhaps the large [ mg / fe ] offsets seen in our bcgs are simply the continuation of this correlation to the largest measured velocity dispersions .
the trend of increasing [ mg / fe ] we see in our three bcgs ( figure 4 ) does agree with their relative velocity dispersions .
a dissipative process is most likely to give rise to these differences , although mergers as a trigger for such a process can not be ruled out ; for instance zepf and ashman ( 1993 ) interpret the bimodal colour ( and hence metallicity ) distributions for globular clusters in some elliptical galaxies , including the bcgs m87 and ngc 1399 , as evidence for a second , merger induced , epoch of cluster formation .
if we had data for ngc 6166 and ngc 6173 alone , it would be tempting to draw a clear distinction between the two galaxies .
ngc 6166 has a rising velocity dispersion , and enhanced [ mg / fe ] , a high globular cluster specific frequency ( bridges _ et al .
_ 1996 ) and an x ray cooling flow .
ngc 6173 has none of these things , yet is a galaxy with a similar luminosity and surface brightness profile ( carter 1977 ) .
ngc 6086 appears to have a mixture of properties : it has a high [ mg / fe ] , yet does not have a rising dispersion profile .
it is important to establish which of these properties depend upon the environment of a galaxy , and similar observations of galaxies with cd morphology in low density environments ( such as ngc 4839 ; oemler 1976 ) are required .
it is not clear how to interpret the rising velocity dispersion profile that we observe in ngc 6166 .
one would expect a rising dispersion profile under any model of bcg formation , since under simple energy arguments the bcg dispersion profile should eventually join that of the cluster as a whole ( as is the case for ngc 1399/fornax ; see section 1 ) . however , we note that the three bcgs with increasing dispersions , and where x ray data for the cluster exist ( ngc 6166 , ic 1101 , abell 3266 ) are all in rich clusters with smooth x - ray morphology . there is no evidence for substructure or peculiar bcg velocities in any of these clusters .
we suggest that these are old , relaxed clusters in which the bcg is at the centre of the cluster potential ; these are the sorts of clusters in which it will be most profitable to search for bcgs with increasing velocity dispersions , in order to study the transition between the bcg and the cluster .
we thank hans - walter rix for the use of his program for absorption line profile analysis , and for extensive advice on this subject .
we would also like to thank nicolas cardiel , javier gorgas , alfonso aragn - salamanca , reynier peletier , jim schombert and an anonymous referee for comments on various aspects of this paper .
thanks also to cardiel , gorgas , and aragn - salamanca for sharing their data with us prior to publication .
the isaac newton telescope is operated on the island of la palma by the royal greenwich observatory in the spanish observatorio del roque de los muchachos of the instituto de astrofisica de canarias .
allen s.w .
, fabian a.c . , 1997 , mnras , 286 , 583 . arnaboldi m. , freeman k.c . , hui x. , capaccioli m. , 1994 , eso messenger , 76 , 40
barnes j.e . , 1996 , in _ proceedings of iau symposium 171 _ , eds r. bender & r.l .
davies , dordrecht : kluwer academic publishers , p. 191 .
bender r. , 1996 , in _ proceedings of iau symposium 171 _ , eds r. bender & r.l .
davies , dordrecht : kluwer academic publishers , p. 181 .
bridges t.j .
, carter d. , harris w.e .
, pritchet c.j . , 1996 , mnras , 281 , 1290 .
bruzual g. , charlot s. , 1997 , in preparation .
buote d.a .
, canizares c.r . , 1996 ,
apj , 457 , 565 .
burstein d. , davies r.l .
, dressler a. , faber s.m . ,
stone r.s . , lynden - bell d. , terlevich r.j . , wegner g. , 1987 , apjs , 64 , 601
. cardiel n. , gorgas j. , aragon - salamanca a. , 1998 , mnras , 298 , 977 . .
carlberg r.g . , 1984a , apj , 286 , 403 .
carlberg r.g . , 1984b , apj , 286 , 416 .
carollo c.m .
, de zeeuw p.t . , van der marel r.p . ,
danzinger i.j . ,
qian e.e . , 1995 ,
apj , 441 , l25 .
carter d. , 1977 , mnras , 178 , 137 .
carter d. , metcalfe n. , 1980 , mnras , 191 , 325 .
carter d. , efstathiou g. , ellis r.s .
, inglis i. , godwin j.g . , 1981 , mnras , 195 , 15p .
carter d. , inglis i. , efstathiou g. , ellis r.s .
, godwin j.g . , 1985 , mnras , 212 , 417 .
carter d. , hau g.k.t .
, thomson r.c . , 1998 , mnras , 294 , 182 .
cohen j. , ryzhov a. , 1997 , apj , 486 , 230 .
couture j. , hardy e. , 1988 , aj , 96 , 867 .
davidge t.j .
, 1992 , aj , 103 , 1512 .
davies r.l . , 1996 , in _ proceedings of iau symposium 171 _ , eds r. bender & r.l .
davies , dordrecht : kluwer academic publishers , p. 37 .
davies r.l . , sadler e.m . ,
peletier r.f . , 1993 , mnras , 262 , 650 .
dehnen , w. , 1995 , mnras , 274 , 919 .
dixon k.l .
, godwin j.g . , peach j.v . , 1989 , mnras , 239 , 539 .
dressler a. , 1979 , apj , 231 , 659 . dubinski j. , 1998 , ap.j . 502 , 141 .
ellis , r.s . ,
gray , p.m. , carter , d. , & godwin , j.g .
, 1984 , mnras 206 , 285 .
faber s.m . ,
friel e.d . , burstein d. , gaskell c.m . , 1985 ,
apjs , 57 , 711 fabian a.c . , 1994 , ara&a , 32 , 277 fabian a.c . , nulsen p.e.j . , 1977 , mnras , 180 , 479 fisher d. , illingworth g. , franx m. , 1995 , apj , 438 , 539 freeman k.c . , 1997 ,
in _ the second stromlo symposium : the nature of elliptical galaxies _ , asp conf .
116 , eds : m. arnaboldi , g.s . da costa , p. saha , pg . 1 garijo a. , athanassoula e. , garcia - gomez c. , 1997 , a&a , 327 , 930 gerhard o.e . , 1993 ,
mnras 265 , 213 gerhard o.e .
, jeske g. , saglia r.p .
, bender r. , 1998 , mnras , 295 , 197 .
gonzalez j.j .
, gorgas j. , 1996 , in _ fresh views of elliptical galaxies _ , asp conf .
86 , p. 225
gonzalez j.j .
, gorgas j. , 1997 , apjl , 481 , l19 gonzalez , j.j .
, gorgas j. , 1998 , in preparation gorgas j. , efstathiou g. , aragon - salamanca a. , 1990 , mnras , 245 , 217 .
gregory s.a . ,
thompson l.a . , 1984 ,
apj , 286 , 422 .
grillmair c.j .
, freeman k.c . ,
bicknell g.v . , carter d. , sommer - larsen j. , couch w.j . , taylor k. , 1994 , apjl , 422 , l9 .
harris w.e . ,
pritchet c.j . , mcclure , r.d .
, 1995 , apj , 441 , 120 .
hau g.k.t . , 1998 ,
thesis , cambridge university .
hau g.k.t .
, carter d. , balcells m. , 1998 , mnras , in press .
hausman m.a .
, ostriker j.p . , 1978 , apj , 224 , 320 .
heckman t.m .
, illingworth g.d .
, miley g.k .
, van breugel w.j.m . , 1985 , apj , 299 , 41 . hernquist l. , 1993 , apj , 409 , 548 .
jones c. , mandel e. , schwarz j. forman w. , murray s.s .
, harnded f.r . , 1979 ,
apjl , 234 , l21 .
kissler - patig m. , brodie j.p . ,
schroder l.l .
, forbes d.a . ,
grillmair c.j . ,
huchra , j.p . , 1998 , aj , 115 , 105 .
larson r.b . , 1975 , mnras , 173 , 761 .
lauer t.r . , 1986 ,
apj , 311 , 34 .
malumuth e.m .
, kirshner r.p . , 1981 , apj , 251 , 508 .
mathews t.a . ,
morgan w.w . , schmidt m. , 1964 , apj , 140 , 35
. mcelroy d.b . , 1995 ,
apjs , 100 , 105 .
mcglynn t.a .
, ostriker j.p . , 1980 , apj , 241 , 915 .
merritt d. , 1984 , apj , 280 , 5 .
merritt d. , 1985 , apj , 289 , 18 .
mihos j.c . , hernquist l. , 1994 , apj , 427 , 112 .
minniti d. , kissler - patig m. , goudfrooij p. , meylan g. , 1998 , aj , 115 , 121 . miralda - escude j. , 1995 , apj , 438 , 514 .
miralda - escude j. , babul a. , 1995 , apj 449 , 18 .
moore , b. , katz , n. , lake , g. , dressler , a. , & oemler , a. , 1996 , nature , 379 , 613 .
mould j. , oke j.b .
, de zeeuw p.t . , nemec j.m . , 1990 , aj , 99 , 1823 . muriel h. , bhringer h. , voges w. , 1996 , in _ proc .
rontgenstrahlung from the universe _ ,
zimmerman , j. trumper , and h. yorke , mpe report 263 , p. 601 .
oegerle w.r . , hill j.m . , fitchett m.j . , 1995 , aj , 110 , 32 .
oemler , a. , 1976 , apj , 209 , 693 .
ostriker j.p . , tremaine s. , 1975 , apj , 202 , l113 .
peletier r.f . , davies r.l . , davis l.e
. , illingworth g.d . , cawson m. , 1990 , aj , 100 , 1091 .
postman m. , lauer t.r . , 1995 , apj , 440 , 28 .
richstone d.o .
, 1976 , apj , 204 , 642 .
rix h .- w .
, white s.d.m . , 1992 , mnras , 254 , 389 . rix h .- w .
, de zeeuw p.t . , cretton n. , van der marel r.p , carollo c. , 1997 , apj , 488 , 702 .
saglia r.p . , bertin g. , bertola f. , danziger j. , dejonghe h. , sadler e.m .
, stiavelli m. , de zeeuw p.t .
, zeilinger w.w . , 1993 ,
apj , 403 , 567 .
schombert j.m . , 1986 , apjs , 60 , 603 .
schombert j.m . , 1987 , apjs , 64 , 643 .
schombert j.m .
, 1988 , apj , 328 , 475 .
sembach k.r .
, tonry , j.l . , 1996 , aj , 112 , 797 .
statler t.s .
, 1991 , apjl , 382 , l11 .
statler t.s .
, smecker - hane t. , cecil g.n . , 1996 , aj , 111 , 1512 .
teague p.f .
, carter d. , gray p.m. , 1990 , apjs , 72 , 715 .
tonry j.l . , 1984 , apj , 279 , 13 .
tonry j.l . , 1985 ,
aj , 90 , 2431 .
trager , s.c . , worthey , g. , faber , s.m . ,
burstein , d. , & gonzalez , j.j . , 1998 , apj suppl . , 116 , 1 . tremaine s. , 1990 , in _ dynamics and interactions of galaxies _ , heidelberg conference , ed .
r. wielen , p. 394 .
van der marel r.p . ,
franx m. , 1993 , apj , 407 , 525 .
van der marel r.p . , rix h .- w . , carter d. , franx m. , white s.d.m .
, de zeeuw p.t . , 1994 , mnras , 268 , 521 .
vazdekis a. , casuso e. , peletier r.f .
, beckman j.e . , 1996 , apjs , 106 , 307 .
white s.d.m . , 1980 , mnras , 191 , 1p .
wilson r.e . , 1953 , carnegie inst .
washington d.c .
601 , _ general catalogue of stellar radial velocities_. worthey g. , 1994 , apjs , 95 , 107 .
worthey g. , faber s.m . , gonzalez j.j .
, 1992 , apj , 398 , 69 .
worthey g. , faber s.m .
, gonzalez j.j . , burstein d. , 1994 , apj suppl , 94 , 687 .
worthey g. , ottaviani d.l .
, 1997 , apjs , 111 , 377 .
zabludoff a.i .
, geller m.j . , huchra j.p . ,
vogeley m.s .
, 1993a , aj , 106 , 1273 .
zabludoff a.i , geller m.j .
, huchra j.p . , ramella m. , 1993b , aj , 106 , 1301 . zepf s. , ashman k.m . , 1993 , mnras , 264 , 611 .
in this appendix the sensitivity of the kinematic parameters to noise for ngc 6166 , ngc 6173 & ngc 6086 is investigated via montecarlo simulations of the effect of both poisson noise and sky residuals .
especially of interest is whether the lowered velocity dispersions measured for points beyond @xmath53 for ngc 6166 and beyond @xmath54 for ngc 6173 are significant . for this purpose
artificial spectra are generated by convolving the optimal stellar templates at the centres of ngc 6166 & ngc 6173 with perfect gaussian losvds .
poisson noise with different random seeds is added to the artificial spectra .
imperfect sky subtraction is investigated by adding blank spectra extracted at large distances from the nucleus .
up to 8 rows are added to mimic the worst case of sky contamination for ngc 6166 , and 16 rows for ngc 6173 .
the kinematic parameters are recovered three times using kinematics , each with different poisson noise seeds and sections of blank sky . for ngc 6166 , losvds with @xmath55 and @xmath56 , 450 , & @xmath57
are modelled .
the artificial spectra are scaled to continuum levels of 12000 , 7200 , 4200 & 2000 counts per pixel , corresponding to distances of 0 , 5 , 8 & @xmath58 from the nucleus respectively . beyond @xmath59 summation
is required to achieve 2000 counts per pixel .
the number of spectra to be summed is 2 , 4 , 5 & 8 for distances of approximately @xmath60 , @xmath61 , @xmath62 and @xmath63 from the nucleus respectively .
the results of the experiments are plotted in fig .
[ fig : n6166monte ] , where the left , centre and right panels correspond to @xmath64 400 , 450 & @xmath57 respectively . for @xmath65
, the mean velocity @xmath17 can be determined to within @xmath66 at distances less than @xmath67 from the nucleus . the scatter increases to @xmath68 at @xmath69 ,
@xmath70 at @xmath71 , and @xmath72 at @xmath73 .
in addition beyond @xmath61 there is a small systematic offset of @xmath74 towards higher velocity . for broader losvds ,
the behavior is similar , except that the recovering accuracy is lower .
if the velocity dispersion is @xmath75 , it can be determined to within @xmath76 at distances less than @xmath67 from the nucleus .
its scatter increases with distance .
in addition , as the contribution of the sky increases with decreasing @xmath77 , the recovered value is systematically lower .
the offset is about @xmath78 at @xmath69 , @xmath79 at @xmath71 , and @xmath80 at @xmath73 .
the behavior is similar if the losvd is broader , except that the magnitude of the systematic offset is larger . at the extreme ,
the recovered value for a gaussian losvd with @xmath81 at @xmath73 is @xmath82 lower . in general , the gauss - hermite moments 3 and 4
can be recovered reasonably well if @xmath83 , to within @xmath84 at distances up to @xmath61 . their scatter increases with @xmath18 .
no obvious systematics are introduced in 3 even at the outermost points , or when @xmath18 is increased to @xmath57 . a small positive offset of order @xmath85 is introduced in 4 for data points beyond @xmath61 , and this tends to increase with @xmath18 . the gauss - hermite moments 5 and 6 can be recovered well to within 0.01 in the inner @xmath59 , and to @xmath86 at @xmath87 .
again , the scatter increases with @xmath18 . in addition a significantly positive offset of @xmath88 is introduced to the @xmath89 at the outer points if sigma is increased beyond @xmath90 .
the conclusions from this exercise are that at distances beyond @xmath91 , where summation of more than 4 spectra are required , systematics start to creep in as the sky contribution increases . at such distances ,
the recovered velocity dispersions are systematically lower .
this shows that we are probably unable to measure losvds broader than @xmath75 beyond these distances , and the low ( @xmath92 ) velocity dispersions beyond @xmath53 actually measured for ngc 6166 are entirely consistent with expectation . for these reasons ,
we measure @xmath17 up to @xmath63 , and @xmath18 , 3 , 4 , h5 & h6 up to @xmath93 from the nucleus . for ngc 6173 , losvds with @xmath94 and @xmath95 , 250 , & @xmath96
are modelled .
the artificial spectra are scaled to continuum levels of 17000 , 3600 & 2000 counts per pixel , corresponding to distances of 0 , 6 , & @xmath97 from the nucleus respectively . beyond @xmath98 summation
is required to achieve 2000 counts per pixel .
the number of spectra to be summed is 4 , 8 & 16 for distances of approximately @xmath99 , @xmath100 and @xmath101 from the nucleus respectively .
the results of the experiments are plotted in fig .
[ fig : n6173monte ] , where the left , centre and right panels correspond to @xmath64200 , 250 & @xmath96 respectively . for @xmath102
, the mean velocity @xmath17 can be determined to within @xmath103 at distances less than @xmath104 from the nucleus . the scatter increases to @xmath68 at @xmath105 and @xmath106 at @xmath107 .
in addition beyond @xmath100 there is a small systematic offset of @xmath108 . for broader losvds ,
the behavior is similar , except that the magnitude of the scatter and the systematic offset is slightly higher .
if the velocity dispersion is @xmath109 , it can be determined to within @xmath110 at distances less than @xmath104 from the nucleus .
its scatter increases with distance .
in addition , as the contribution of the sky increases with decreasing @xmath77 , the recovered value is systematically lower .
the offset is about @xmath111 at @xmath112 , @xmath113 at @xmath114 , and @xmath115 at @xmath107 .
the magnitude of the systematic offset is larger if the losvd is broader , and smaller if the losvd is narrower . at the extreme ,
the recovered value for a gaussian losvd with @xmath116 at @xmath107 is @xmath117 lower . in general , the gauss - hermite moments 3 and 4
can be recovered reasonably well if the @xmath77 is high .
if @xmath118 , they can be determined to within @xmath86 at distances up to @xmath98 . their scatter increases with @xmath18 and sky contribution .
a small positive offset in 3 is introduced with increasing sky contribution .
no significant systematics are introduced in 4 , except for the outermost points at @xmath119 , where 4 tends to be lower .
the gauss - hermite moments 5 and 6 can be recovered well only near the galaxy centre ( @xmath120 ) .
the scatter increases with @xmath18 and increasing sky contribution . if @xmath121 , there are no obvious systematics in 5 with increasing sky contribution .
however , at distances larger than @xmath99 5 is scattered upwards if @xmath102 , and downwards if @xmath116 .
6 tends to be scattered downwards at distances larger than @xmath99 for all @xmath18 .
the tests for ngc 6173 should also be applicable to ngc 6086 , as both galaxies have similar velocity dispersions , and are at similar distances .
for ngc 6086 , the continuum levels are 14600 and 2000 counts per pixel at @xmath122 and @xmath67 . beyond @xmath67 summation
is required .
the number of spectra to be summed is 4 , 6 & 15 for distances of approximately @xmath123 , @xmath59 and @xmath100 from the nucleus respectively .
the conclusions from this exercise are that at distances beyond @xmath124 , where summation of more than 4 spectra are required , systematics start to creep in as the sky contribution increases . at such distances ,
the recovered velocity dispersions are systematically lower , but the problem is not as severe as in ngc 6166 .
this shows that the fall in @xmath18 beyond @xmath54 is probably more gradual in reality .
the significance of the monte carlo tests is that the @xmath18 in ngc 6173 is consistent with being flat or falling beyond @xmath54 , but inconsistent with rising outwards .
we measure @xmath17 , @xmath18 & 3 up to @xmath53 , 4 up to @xmath125 , and 5 & 6 up to @xmath99 from the nucleus .
the tests described above show that the systematic errors on the recovery of @xmath1 caused by noise are small in the case of @xmath1 = 0 .
it is important also to ask what other values of @xmath1 can be recovered by our techniques . for ngc 6166 we model losvds with @xmath126 ans
@xmath65 , and @xmath127 - 0.10 , -0.05 , 0.00 , + 0.05 , + 0.10 and + 0.15 respectively , and the results are plotted in fig .
[ fig : h4test ] .
the systematic bias on @xmath18 at low @xmath77 is marginally worse for positive @xmath1 ( second panel up in each column ) , and the main result of this exercise is that there is a systematic underestimation of @xmath39 for @xmath128 . our measurement of high @xmath1 in ngc 6166 could possibly still be a systematic underestimate .
our kinematic parameters are all determined using optimal best fit templates constructed from 28 stellar spectra . to demonstrate that this procedure does not itself introduce systematic effects into the kinematic parameters , here we repeat the analysis for ngc 6166 using individual template stars .
moreover to demonstrate the effect of template choice on the derived parameters we have carried out this exercise with templates with a wide range of spectral type . in fig .
[ fig : templatetest ] we plot the kinematic parameters for ngc 6166 derived from the five individual templates : hr7555 ( g6iii , open squares ) ; hr4864 ( g7v , filled triangles ) ; hr6674 ( k0iii , crosses ) ; hr4676 ( k4iii , open circles ) and hr8057 ( m1iii , thin rectangles ) .
the trends in the derived parameters are identical to those in fig .
[ fig : hermplot ] , although there are differences in the zero points which depend upon the template choice .
the difference in the velocity zero points is not real , it is because this procedure determines the velocity relative to the first star in the list of templates , if there is only one it depends upon the radial velocity of that star . for @xmath18 and all of the hermite terms
the m1iii star hr8057 gives a very different zero point , the dispersions are so much lower that many of them are off the bottom of the dispersion panel in fig .
[ fig : templatetest ] .
however this star is a very poor fit to the galaxy spectra and does not contribute much to the optimal composite templates .
for the other templates , @xmath1 is tightly defined around + 0.10 , suggesting that template mismatch errors do not contribute to our large positive values of @xmath1 . on the other hand
there are large zero point differences in @xmath34 between the g and k star templates , to get reliable values of @xmath34 it is important that the templates used fit well . the velocity dispersion zero points differ with a total spread of 50 @xmath129 ( again neglecting hr8057 ) , so the uncertainty of the zero point of the velocity dispersion profile due to template mismatch might be estimated at 25 @xmath129 .
in this appendix we describe the procedures for obtaining absorption line indices in the lick system .
line indices are extracted from individual , de - redshifted spectra following the procedures outlined in worthey & ottaviani ( 1997 ) , adopting the passband definitions of worthey _ et al . _
the blue - ward continuum bandpass for 2 g is redefined to be [email protected] instead of [email protected] , because the spectra of the template stars only start from about 4931 .
the difference between the mg index 2g@xmath130 under this alternative definition , and the lick 2 g index is tiny ( see below ) .
the molecular band 2g@xmath130 is measured in pseudo magnitudes , whilst the atomic absorption line indices 52 and 53e are measured as equivalent widths in angstroms . lccc feature & type & central bandpass & continuum bandpasses + & & ( ) & ( ) + + 2g@xmath130&molecular band & 5154.1255196.625 & 4931.6254957.625 , 5301.1255366.125 + 2 g & molecular band & 5154.1255196.625 & 4895.1254957.625 , 5301.1255366.125 + 52 & atomic absorption line & 5245.6505285.650 & 5233.1505248.150 , 5285.6505318.150 + 53e & atomic absorption line & 5312.1255352.125 & 5304.6255315.875 , 5353.3755363.375 + first , our spectra are broadened to a resolution of @xmath131 fwhm , in common with the lick / ids resolution in this part of the spectrum . in ngc 6166
, there is significant [ ni ] emission at @xmath132 .
as the [ ni ] emission line is halfway inside the 2 g passband , its flux can artificially lower the 2 g by @xmath133 at the nucleus .
therefore the flux contribution between rest wavelengths 5183.125 @xmath134 and 5196.526 @xmath134 is measured from the residuals of the kinematic fitting and subtracted off the flux in the 2 g passband .
the indices are then corrected to a zero dispersion system by applying correction factors .
these are estimated by broadening the optimal stellar template of ngc 6166 to different velocity dispersions , and are summarized in table [ tab : n6166dispcor ] .
the correction for 2 g is insignificant even at large dispersions , whilst 53e requires a rather large correction .
these corrections are not sensitive to the choice of template used to determine these corrections , in the worst case ( 53e at high velocity dispersion ) the total range in the factor determined using a range of templates is 0.034 , implying an error in the factor as determined from the best fit template of at most 0.01 .
cccc @xmath18 & 2 g & 52 & 53e + ( @xmath129 ) & & & + + 160 & 0.998 & 0.938 & 0.895 + 180 & 0.998 & 0.924 & 0.868 + 200 & 0.997 & 0.909 & 0.842 + 220 & 0.997 & 0.895 & 0.815 + 240 & 0.997 & 0.879 & 0.785 + 260 & 0.996 & 0.864 & 0.756 + 280 & 0.996 & 0.849 & 0.727 + 300 & 0.995 & 0.834 & 0.697 + 320 & 0.995 & 0.819 & 0.668 + 340 & 0.994 & 0.804 & 0.640 + 360 & 0.993 & 0.788 & 0.612 + 380 & 0.991 & 0.773 & 0.586 + 400 & 0.990 & 0.758 & 0.560 + 420 & 0.988 & 0.742 & 0.536 + 440 & 0.987 & 0.726 & 0.512 + to convert 2g@xmath130 to 2 g , a scaling relationship is required .
this is obtained by comparing the 2 g indices for 9 template stars measured under the lick and our alternative definition .
the template stars were observed on the aat in march 1996 using the rgo spectrograph , and their spectra span the wavelength range 4850 5610 @xmath134 , and are broadened to @xmath131 fwhm resolution .
the fact that these spectra are taken with a completely different instrument and telescope does not pose a problem , as the only variable is the change in bandpass definition .
the 2 g indices under worthey s definition are plotted against those under the alternative definition in fig .
[ fig : mglickvsalt ] , and the following relationship is obtained by a linear least - squares fit : @xmath135 because the changes introduced by redefining the continuum bandpass are only 0.5% for 2 g , they are unlikely to be a major source of error .
because our spectra are not flux calibrated , there might be a zero point difference between our measured indices and the lick values , due to the curvature of the response function .
of our standard stars , only one , hr 5277 has been measured on the lick system .
worthey ( private communication ) ascribes low weight to the observations of this particular star , and in any case using observations of a standard star to determine the zero points for galaxies is subject to systematic error , due to the substantial redshifts of our galaxies , and the corresponding difference in the grating settings for the observations .
thus we are unable to tie our measurements to the lick system directly . on the other hand cardiel _
( 1997 ) have measurements which we , and they , believe to be on the lick system .
figure [ fig : uscardiel ] demonstrates good agreement between the 2 g values in the region in which the data of cardiel _ et al .
_ are reliable , and thus we believe that the zero point offset between our values and the lick system is small , at least in 2 g .
( 1998 ) publish lick indices for a 1.4 x 4 arcsecond aperture centred on the nucleii of ngc 6086 and ngc 6166 .
for ngc 6086 they find @xmath136 ; @xmath137 ; @xmath138 . for ngc 6166
they find @xmath139 ; @xmath140 ; @xmath141 . these are consistent with our results , although the error bars on their iron line equivalent widths are disappointingly large .
although some uncertainty remains as to the zero point of the iron line indices , it seems that our indices are within 0.01 magnitudes of the lick system in the case of 2 g , and within 0.3 angstrom in the case of the combined iron line index . | we present kinematic parameters and absorption line strengths for three brightest cluster galaxies , ngc 6166 , ngc 6173 and ngc 6086 .
we find that ngc 6166 has a velocity dispersion profile which rises beyond 20 arcsec from the nucleus , with a halo velocity dispersion in excess of 400 km s@xmath0 .
all three galaxies show a positive and constant @xmath1 hermite moment .
the rising velocity dispersion profile in ngc 6166 thus indicates an increasing mass - to - light ratio .
rotation is low in all three galaxies , and ngc 6173 and ngc 6086 show possible kinematically decoupled cores .
all three galaxies have 2 g gradients similar to those found in normal bright ellipticals , which are not steep enough to support simple dissipative collapse models , but these could be accompanied by dissipationless mergers which would tend to dilute the abundance gradients .
the [ mg / fe ] ratios in ngc 6166 and ngc 6086 are higher than that in ngc 6173 , and if ngc 6173 is typical of normal bright ellipticals , this suggests that cds can not form from late mergers of normal galaxies . # 1#1 0@xmath2 52fe@xmath3 53efe@xmath4 54efe@xmath5 50fe@xmath6 2gmg@xmath7 2mg@xmath7
1mg@xmath8 galaxies : normal galaxies : elliptical and lenticular , cd |
bridge number is a classical link invariant originally introduced by schubert as a tool to study companion tori . in @xcite
, schubert proves the remarkable fact that given a composite knot @xmath0 with summands @xmath1 and @xmath2 , the following equality holds : @xmath3 , where @xmath4 denotes the bridge number of a link @xmath5 .
schultens gives a modern proof of this equality in @xcite .
connected sum is a classical and intentionally restrictive method of amalgamating two links in @xmath6 together to create a new link in @xmath6 .
tangle products are the natural generalization of this amalgamation operation . roughly speaking , to form an @xmath7-strand tangle product of links @xmath8 and @xmath9 remove an @xmath7-strand rational tangle from the 3-sphere containing @xmath8 and the three sphere containing @xmath9 .
now , glue the resulting tangles together via some homeomorphism of the @xmath10-punctured sphere .
the result is a tangle product , denoted @xmath11 . for a rigorous definition see section [ sec : tangleprod ] . in particular , connected sums are 1-strand tangle products and conway products are 2-strand tangle products .
conway products were studied in @xcite where scharlemann and tomova produced conway products which respected multiple bridge surfaces . how bridge number behaves with respect to conway products
was studied in @xcite .
the goal of this paper is to generalize schubert s equality for bridge number to the operation of tangle product .
because of their generality and the choices involved , tangle products are exceptionally poorly behaved .
for example , for any two @xmath7 bridge knots @xmath8 and @xmath9 there exists an @xmath7-strand tangle product @xmath11 isotopic to the unknot .
hence , we will have to restrict our hypothesis to achieve a meaningful lower bound for @xmath12 in terms of @xmath13 and @xmath14 .
similarly , if @xmath15 is the unknot , there exist tangle products @xmath16 of arbitrarily high bridge number .
this observation implies , in the absence of additional information , that there does not exist an upper bound for @xmath12 in terms of @xmath13 and @xmath17 .
the following is the main theorem .
[ main ] given an @xmath7-strand tangle product @xmath18 such that there exists a minimal bridge sphere for @xmath18 of distance at least three and the product sphere @xmath19 is c - incompressible , then @xmath20 . in @xcite , the key additional hypothesis needed to produce a lower bound on the bridge number of a conway product in terms of the bridge number of the two factor links was that bridge position and thin position
coincide for the conway product .
in contrast , the result presented here is heavily dependent on the hypothesis that the tangle product has a minimal bridge sphere of distance at least three .
this hypothesis allows for a much stronger structure theorem then that found in @xcite and , thus , a more restrictive lower bound on the bridge number of a tangle product .
it is important to note that having the property that bridge position is thin position and having the property that a minimal bridge sphere is distance at least three are believed to be independent conditions .
the following structure theorem is of independent interest and is proven in section [ structure ] , h - level and taut are defined in section [ prelim ] , and standard and adjacent are defined in section [ structure ] .
[ thm : struc ] suppose there exists a bridge sphere @xmath21 for @xmath0 of distance three or greater and @xmath19 is taut with respect to an @xmath22-level embedding of @xmath21 , then the following hold : 1)there do not exist standard saddles @xmath23 , @xmath24 , and @xmath25 for @xmath19 such that @xmath23 is adjacent to @xmath24 and @xmath24 is adjacent to @xmath25 .
2)every outermost disk of the foliation of @xmath19 induced by the standard height function contains at least one point of @xmath26 .
bachman and schleimer showed that twice the genus plus the number of boundary components of of an essential surface serves as an upper bound for the distance of any bridge sphere for a knot @xcite .
in other words , a high distance bridge sphere forces a high _ intrinsic _ complexity for any essential embedded surface in a knot complement . in contrast , theorem [ thm : struc ] implies that if @xmath0 has a bridge sphere of distance 3 or greater , then any essential punctured sphere can be isotoped so that the number of saddles in the induced foliation is bounded with respect to the number of punctures .
colloquially , if there exists a bridge sphere that is not low distance , then every essential punctured sphere has low _ extrinsic _ complexity .
in this paper we study smooth links in @xmath6 .
our central tool will be the standard height function on @xmath6 , @xmath27 $ ] .
the level surfaces of @xmath22 foliate @xmath6 into concentric 2-spheres and two exceptional points .
the * bridge number * of a link is the minimal number of maxima of @xmath28 over all isotopic morse embeddings of @xmath0 .
a * tangle * is an ordered pair @xmath29 where @xmath30 is a 3-ball and @xmath31 is a properly embedded collection of arcs and loops .
an * untangle * is a tangle @xmath29 such that @xmath32 is a collection of boundary parallel arcs .
we say an embedded surface in @xmath6 is @xmath33-punctured if it meets @xmath0 transversely in @xmath33 points .
a * bridge sphere * for a link @xmath0 is an @xmath33-punctured sphere decomposing @xmath34 into two trivial tangles @xmath35 and @xmath36 .
a bridge sphere @xmath21 is * h - level * if there exists a regular value @xmath37 such that @xmath21 is isotopic to @xmath38 . given an embedded morse surface @xmath19 in @xmath6 , let @xmath39 be the singular foliation on @xmath19 induced by @xmath40 .
a * saddle * is any leaf of this foliation homeomorphic to the wedge of two circles . by standard position
, we can assume that all saddles of @xmath39 are disjoint from @xmath0 .
[ pairbridge]let @xmath0 be a link embedded in @xmath6 and @xmath19 be a surface embedded in @xmath6 which meets @xmath0 transversely .
we say that the pair @xmath41 is in bridge position with respect to the standard height function on @xmath6 if @xmath22 is a morse function when restricted to both @xmath0 and @xmath19 and there exist @xmath42 $ ] such that : 1 .
all maxima of @xmath0 and all maxima of @xmath19 lie in @xmath43 2 .
all minima of @xmath0 and all minima of @xmath19 lie in @xmath44 3 .
all saddles of @xmath19 and all intersection points @xmath45 lie in @xmath46 [ pairbridgeiso ] for any embeddings of @xmath0 and @xmath19 in @xmath6 there is an isotopy of first @xmath19 and subsequently @xmath0 such that the resulting pair @xmath41 is in bridge position .
moreover , these isotopies fix @xmath19 and @xmath0 outside of a neighborhood of their maxima and minima , preserve the number of maxima of @xmath47 , and the number of saddles of @xmath48 . after a small isotopy
, we can assume that @xmath49 and @xmath50 are morse functions and @xmath19 intersects @xmath0 transversely .
additionally , by general position , we can assume that both @xmath0 and @xmath19 are disjoint from both @xmath51 and @xmath52 .
let @xmath53 be the largest value among the heights of all saddles of @xmath48 and all points of @xmath45 .
let @xmath54 be the smallest value among the heights of all saddles of @xmath48 and all points of @xmath45 .
let @xmath55 and @xmath56 .
let @xmath57 be the highest maximum of @xmath19 that lies below @xmath58 .
let @xmath59 be a monotone arc connecting @xmath57 to any point in @xmath6 above @xmath58 .
since @xmath57 is the highest maximum below @xmath58 , we can choose @xmath59 so that the interior of @xmath59 is disjoint from both @xmath0 and @xmath19 .
the portion of the boundary of a regular neighborhood of @xmath59 lying above @xmath57 is a monotone disk @xmath60 such that @xmath60 is disjoint from @xmath0 and @xmath19 except in its boundary .
let @xmath61 be a regular neighborhood of @xmath57 in @xmath19 .
@xmath60 together with @xmath61 co - bound a 3-ball whose intersection with @xmath19 is @xmath61 .
isotope @xmath61 to @xmath60 across this 3-ball .
this isotopy fixes @xmath0 , is supported in a neighborhood of @xmath57 in @xmath19 , and raises one maximum of @xmath19 above @xmath58 .
repeat this process until all maxima of @xmath19 lie above @xmath58 .
now that there are no maxima of @xmath19 between @xmath62 and @xmath58 we can similarly isotope all the maxima of @xmath63 above @xmath58 via an isotopy that fixes @xmath19 and is supported on a neighborhood of the maxima of @xmath64 .
hence , we have achieved @xmath65 in the definition of @xmath41 bridge position while preserving the number of maxima of @xmath0 and the number of saddles of @xmath19 . by a symmetric argument
, we can also achieve @xmath66 in the definition of @xmath41 bridge position while preserving the number of maxima of @xmath0 and the number of saddles of @xmath19 .
after these isotopies , our choice of @xmath67 and @xmath68 guarantee that @xmath69 in the definition of @xmath41 bridge position is satisfied .
a punctured surface @xmath19 is * taut * with respect to an @xmath22-level bridge sphere @xmath21 if @xmath39 contains the fewest number of saddles subject to @xmath41 being in bridge position .
this notion of taut is different then that found in @xcite . specifically , a taut surface in @xcite is one that has a minimal number of saddles subject to @xmath70 having a minimal number of maxima . in this paper ,
a taut surface has a minimal number of saddles subject to a specified bridge sphere appearing as a level sphere of the height function .
in this section we use the notion of taut introduced in the previous section to develop constraints on @xmath48 when @xmath19 is an embedded essential punctured sphere . for a more detailed discussion of the following definitions and their applications see @xcite .
any given saddle @xmath71 , lies in a level sphere @xmath72 .
let @xmath73 be the closure of the component of @xmath74 that is disjoint from @xmath75 and @xmath76 be the closure of the component of @xmath77 that is disjoint from @xmath78 .
a subdisk @xmath60 in @xmath39 is monotone if its boundary is entirely contained in a leaf of @xmath39 and the interior of @xmath60 is disjoint from every saddle in @xmath39 . in practice , we will use the term subdisk in a slightly broader sense , allowing @xmath79 to be immersed in @xmath19 ( i.e. @xmath79 is a saddle ) .
we say a monotone disk is * outermost * if its boundary is @xmath80 for some saddle @xmath23 and label the disk @xmath81 . similarly ,
if @xmath80 bounds an outermost disk @xmath81 , we say @xmath23 is an outermost saddle .
it is usually the case that only one of @xmath78 and @xmath75 is the boundary of an outermost disk , so , our convention is to relabel so that @xmath82 .
we say @xmath23 is an * inessential saddle * if @xmath23 is an outermost saddle and @xmath81 is disjoint from @xmath0 .
suppose @xmath23 is an outermost saddle .
@xmath83 cuts @xmath84 into two 3-balls .
the one that contains @xmath81 is again cut by @xmath81 into two 3-balls @xmath85 and @xmath86 .
we chose the labeling of @xmath85 and @xmath86 so that @xmath87 .
we say a saddle @xmath23 is * standard * if there is a monotone disk @xmath88 in @xmath19 such that @xmath89 and @xmath88 is disjoint from @xmath0 . by general position arguments
, we can assume every saddle @xmath23 in @xmath39 has a bicollared neighborhood in @xmath19 that is disjoint from @xmath0 and all other singular leaves of @xmath39 .
the boundary of this bicollared neighborhood consists of three circles @xmath90 , @xmath91 , and @xmath92 where @xmath90 and @xmath91 are parallel to @xmath78 and @xmath75 respectively .
we can assume @xmath90 , @xmath91 , and @xmath92 are level with respect to @xmath22 and that @xmath90 and @xmath91 lie in the same level surface .
figure 1 illustrates all of the terminology outlined above .
for the remainder of the paper @xmath19 will always denote a punctured sphere .
we will denote the point @xmath51 as @xmath93 and the point @xmath52 as @xmath94 [ inftyball]let @xmath23 be an outermost saddle in @xmath39 .
there is an ambient isotopy of @xmath19 that fixes @xmath0 , lowers minima of @xmath19 , raises maxima of @xmath19 , and fixes @xmath19 outside of a neighborhood of the maxima and minima of @xmath19 such that , after this isotopy , @xmath85 does not contain @xmath93 or @xmath94 .
see lemma 1 @xcite .
the following lemma is an extension of lemma 2 @xcite to our alternative notion of taut .
[ iness]suppose @xmath21 is an @xmath22-level bridge sphere for a link @xmath0 .
if @xmath48 contains an inessential saddle , then @xmath19 is not taut with respect to @xmath21 .
suppose @xmath41 is in bridge position with respect to @xmath21 , an @xmath22-level bridge sphere .
let @xmath23 be an inessential saddle in @xmath48 .
we can assume @xmath81 contains a unique maximum and , by lemma [ inftyball ] , @xmath85 does not contain @xmath93 .
let @xmath95 be the interval in the definition of @xmath41 bridge position .
there exists an open interval @xmath96 such that @xmath97 and @xmath98 contains no saddles , maxima or minima of @xmath48 , no maxima or minima of @xmath0 and no points of @xmath26 other than @xmath23 .
let @xmath99 . horizontally shrink and vertically lower @xmath85 so that the result of the isotopy , call it @xmath100 is contained in @xmath101 .
this isotopy preserves the saddle structure of @xmath48 and the number of critical values of @xmath64 .
let @xmath102 be the image of @xmath19 under this isotopy .
similarly , let @xmath103 be the image of @xmath81 under this isotopy and let @xmath104 be the unique maximum of @xmath103 .
let @xmath105 be the level surface containing @xmath104 .
hence , @xmath106 consists of the point @xmath104 and a collection of circles .
one such circle @xmath107 is parallel in @xmath102 to @xmath108 .
since the region between @xmath105 and @xmath83 outside of @xmath100 meets @xmath19 and @xmath0 in a collection of monotone annuli and vertical arcs respectively , we can choose a point @xmath7 in @xmath107 and an arc @xmath59 in @xmath105 that is disjoint from @xmath102 except at its boundary @xmath109 .
additionally , let @xmath110 be an arc in @xmath102 that does not meet @xmath0 , has boundary @xmath109 and is transverse to @xmath111 everywhere accept where it passes through @xmath112 so that @xmath59 and @xmath110 cobound a vertical disk @xmath113 that is disjoint from @xmath0 and disjoint from @xmath19 except along @xmath110 .
isotope @xmath102 along @xmath113 to effectively cancel a saddle with a maximum .
see figure [ fig : thm1fig1.eps ] .
let @xmath114 be the image of @xmath102 under this second isotopy .
notice that the ambient isotopies described fix all of @xmath6 below @xmath115 , hence , @xmath115 remains a bridge sphere for @xmath0 isotopic to @xmath21 .
the only way for @xmath116 to fail to be in bridge position is for maxima of @xmath114 or @xmath0 to lie below a saddle of @xmath114 or a point of @xmath26 . in this case
use the isotopy from lemma [ pairbridgeiso ] to raise the maxima of @xmath114 and subsequently the maxima of @xmath0 .
thus , we have produced an ambient isotopy of @xmath19 and @xmath0 which reduces the number of saddles of @xmath19 but preserves both @xmath21 as an @xmath22-level bridge sphere and @xmath41 in bridge position . hence @xmath19 is not taut with respect to @xmath21 .
we say @xmath23 is a * removable saddle * if @xmath23 is an outermost saddle where @xmath81 has a unique maximum(minimum ) and @xmath117 has a local end - point maximum(minimum ) at every point of @xmath118 .
see figure [ fig : removeable.eps ] .
otherwise , we say @xmath23 is non - removable .
the following is an adaptation of lemma 3 of @xcite to the notion of taut presented here .
[ remove]suppose @xmath21 is an @xmath22-level bridge sphere for a link @xmath0 .
if @xmath48 contains a removable saddle , then @xmath19 is not taut with respect to @xmath21 .
suppose @xmath41 is in bridge position with respect to @xmath21 , an @xmath22-level bridge sphere .
let @xmath23 be an removable saddle in @xmath48 .
we can assume @xmath81 contains a unique maximum and , by lemma [ inftyball ] , @xmath85 does not contain @xmath93 . applying the isotopy presented in lemma [ iness ]
, we see that each point @xmath119 together with the image of @xmath120 in @xmath100 bound monotone subarcs of @xmath121 .
see figure [ fig : tightcase2.eps ] . since @xmath23 is removable neither the @xmath120 nor
the @xmath122 is a maximum or minimum .
thus , we have eliminated a saddle of @xmath19 while preserving @xmath41 bridge position , a contradiction to tautness of @xmath19 .
@xmath19 decomposes @xmath84 into two 3-balls @xmath123 and @xmath124 .
let @xmath23 be a saddle in @xmath39 and @xmath5 be the level sphere containing @xmath90 and @xmath91 .
@xmath125 is composed of two disks and an annulus @xmath126 .
if a collar of @xmath127 in @xmath126 is contained in @xmath123 , then we say @xmath23 is * unnested * with respect to @xmath123 .
if not , we say @xmath23 is nested with respect to @xmath123 .
we define nested and unnested with respect to @xmath124 similarly .
note that nested with respect to @xmath123 is the same as unnested with respect to @xmath124 and nested with respect to @xmath124 is unnested with respect to @xmath123 .
two saddles @xmath128 and @xmath129 in @xmath39 are * adjacent * if , up to labeling , @xmath80 and @xmath130 cobound a monotone annulus in @xmath19 that is disjoint from @xmath0 .
the following lemma is an extension of lemma 3 in @xcite to our alternative notion of taut .
[ nesting ] suppose @xmath21 is an @xmath22-level bridge sphere for a link @xmath0 .
if @xmath48 contains adjacent saddles @xmath23 and @xmath25 where @xmath23 is a standard saddle and @xmath23 and @xmath25 are nested with respect to different 3-balls , then @xmath19 is not taut with respect to @xmath21 .
do to the symmetry of the argument we can assume that @xmath88 has a unique maximum .
since @xmath23 and @xmath25 are adjacent , then , up to relabeling , @xmath131 and @xmath132 cobound a monotone annulus @xmath126 in @xmath19 that is disjoint from @xmath0 . after a small isotopy of @xmath0 and @xmath19
, we can assume that @xmath133 meets @xmath134 in a collection of points and simple closed curves .
after a small tilt of @xmath134 , @xmath135 is a monotone disk .
eliminate the saddle @xmath25 by applying the isotopy from the proof of lemma [ iness ] to the 3-ball cobounded by the monotone disk @xmath135 and the level disk @xmath136 that is disjoint from @xmath137 .
see figure [ fig : nesting.eps ] . as in the proof of lemma
[ remove ] , we see that each point of @xmath138 together with its image under the isotopy cobound monotone sub arcs of @xmath0 .
similarly , each curve @xmath139 together with its image under the isotopy cobounds monotone annulus in @xmath19 .
thus , we have produced an ambient isotopy of @xmath19 and @xmath0 which reduces the number of saddles of @xmath19 but preserves both @xmath21 as an @xmath22-level bridge sphere and @xmath41 in bridge position .
hence , @xmath19 is not taut with respect to @xmath21 .
the previous lemmas in this section are independent of bridge sphere distance .
below we define the distance of a bridge sphere and obtain additional constraints on taut punctured spheres .
let @xmath21 be a @xmath10-punctured bridge sphere separating @xmath34 into two trivial @xmath7-strand tangles @xmath35 and @xmath36 .
let @xmath140 be the curve complex for @xmath21 .
let @xmath141 be the set of all isotopy classes of essential simple closed curves in @xmath142 that bound disks in @xmath143 .
define @xmath144 analogously .
the * distance * of @xmath21 , denoted @xmath145 , is the distance between @xmath141 and @xmath144 in @xmath140 where the metric structure of @xmath140 arises from assigning a length of one to each edge .
[ prime ] suppose @xmath0 is a link with a bridge sphere @xmath21 of distance 3 or greater , then @xmath0 is not split and @xmath0 is prime .
this follows immediately from theorem 5.1 of @xcite .
[ triple ] suppose there exists a bridge sphere @xmath21 for @xmath0 of distance three or greater .
if @xmath19 is c - incompressible punctured sphere taut with respect to an @xmath22-level embedding of @xmath21 , then there do not exist standard saddles @xmath23 , @xmath24 , and @xmath25 in @xmath48 such that @xmath23 is adjacent to @xmath24 and @xmath24 is adjacent to @xmath25 . by lemma [ prime ] ,
we know @xmath0 is non - split and prime .
we will proceed by proving the contrapositive of the above statement .
suppose that three such saddles @xmath23 , @xmath24 and @xmath25 do exist .
if @xmath48 contains inessential or removable saddles , then @xmath19 is not taut by lemma [ iness ] or lemma [ remove ] . if @xmath23 and @xmath24 are nested with respect to different three balls , then @xmath19 is not taut by lemma [ nesting ] .
similarly , if @xmath24 and @xmath25 are nested with respect to different three balls then @xmath19 is not taut by lemma [ nesting ] .
hence , we can assume that all three saddles @xmath23 , @xmath24 and @xmath25 have a common nesting and @xmath48 contains no inessential and no removable saddles .
assume that @xmath146 has a unique maximum and @xmath147 .
if @xmath146 has a unique minimum or @xmath148 the proof follows similarly . by lemma [ pairbridgeiso ] , we can assume there is a partitioning of the critical values of @xmath28 and @xmath50 as in definition [ pairbridge ] .
by standard morse theory arguments , we can assume that the collection @xmath149 of all critical points of @xmath28 , all critical points of @xmath50 , and all points of @xmath45 occur at distinct heights .
choose @xmath150 so that @xmath150 is strictly between @xmath151 and the height of the next higher element of @xmath149 .
additionally , @xmath152 is a bridge sphere for @xmath0 that is isotopic to @xmath21 .
although an abuse of notation , we will refer to @xmath152 as @xmath21 .
recall the definition of @xmath153 and @xmath154 from section 3 and figure [ fig : labels1.eps ] .
we can assume that @xmath153 and @xmath154 lie on @xmath21 .
let @xmath155 be the monotone annulus with boundary @xmath156 .
since @xmath23 is assumed to be higher than @xmath25 then @xmath155 intersects @xmath157 in a single simple closed curve @xmath158 .
in particular , @xmath158 is isotopic to both @xmath159 and @xmath160 .
see figure [ fig : triple.eps ] .
claim : if any of @xmath153 , @xmath154 or @xmath158 bounds a zero or once punctured disk in @xmath21 , then @xmath19 is not taut .
proof of claim : since all three saddles @xmath23 , @xmath24 and @xmath25 have a common nesting , each of @xmath153 , @xmath154 and @xmath158 bound pairwise disjoint disks @xmath161 , @xmath162 and @xmath163 respectively in @xmath21 . to prove the claim it suffices to show that each of the disks @xmath161 , @xmath162 and @xmath163 meet @xmath0 in at least two points .
suppose , to form a contradiction , that @xmath161 is disjoint from @xmath0 .
@xmath164 is a collection of disjoint simple closed curves .
an innermost such curve @xmath165 bounds a disk @xmath60 in @xmath21 that is disjoint from @xmath0 and meets @xmath19 only in its boundary . by c
- incompressibility of @xmath19 , @xmath165 also bounds a disk @xmath166 in @xmath19 that is disjoint from @xmath0 . since @xmath0 is non - split , @xmath60 and @xmath166 cobound a 3-ball disjoint from @xmath0 .
hence , we can eliminate @xmath165 as an curve of intersection by isotopying @xmath166 across this 3-ball and just past @xmath60 .
this isotopy leaves @xmath0 fixed and can only decrease the number of saddles in @xmath111 . by repeating this process , we can eliminate all curves of intersection of @xmath19 with @xmath167 .
again , by c - incompressibility of @xmath19 , @xmath168 bounds a disk @xmath166 in @xmath169 which is disjoint from @xmath0 . since @xmath0 is non - split @xmath166 and @xmath161 cobound a 3-ball .
isotope @xmath166 across this 3-ball to @xmath161 while fixing @xmath0 .
if @xmath166 contains @xmath23 , then this isotopy eliminates @xmath23 , @xmath24 and @xmath25 while creating no new saddles . in this case
@xmath19 was not taut .
if @xmath166 does not contain @xmath23 then , after the isotopy , @xmath23 is an inessential saddle .
hence , @xmath19 is not taut , by lemma [ iness ] .
suppose , to from a contradiction , that @xmath161 meets @xmath0 in exactly one point .
@xmath164 is a collection of disjoint simple closed curves .
an innermost such curve @xmath165 bounds a disk @xmath60 in @xmath21 that meets @xmath19 only in its boundary . if @xmath60 is disjoint from @xmath0 then apply the argument in the preceding paragraph to eliminate @xmath165 .
hence we can assume @xmath60 meets @xmath0 exactly once .
by c - incompressibility of @xmath19 , @xmath165 also bounds a disk @xmath166 in @xmath19 that meets @xmath0 exactly once . since @xmath0 is prime
, @xmath60 and @xmath166 cobound a 3-ball containing an unknotted arc of @xmath0 .
hence , we can eliminate @xmath165 as an curve of intersection isotopying @xmath170 across this 3-ball and just past @xmath171 .
this isotopy leaves @xmath0 fixed and can only decrease the number of saddles in @xmath111 . by repeating this process , we can eliminate all curves of intersection of @xmath19 with @xmath167 .
again , by c - incompressibility of @xmath19 , @xmath168 bounds a disk @xmath166 in @xmath169 which meets @xmath0 exactly once and @xmath172 is the boundary of a three ball containing an unknotted arc of @xmath0 .
isotope @xmath166 across this 3-ball to @xmath161 while fixing @xmath0 .
if @xmath166 contains @xmath23 , then this isotopy eliminates @xmath23 , @xmath24 and @xmath25 while creating no new saddles . in this case
@xmath19 was not taut .
if @xmath166 does not contain @xmath23 then , after the isotopy , @xmath23 is a removable saddle .
hence , @xmath19 is not taut , by lemma [ remove ] . by applying nearly identical arguments
, we can show that both @xmath162 and @xmath163 meets @xmath0 in at least two points .
the claim then follows.@xmath173 let @xmath105 be the three ball above @xmath21 and @xmath174 be the three ball below @xmath21 . by construction @xmath153 and @xmath154 cobound an annulus @xmath175 properly embedded in @xmath174 and disjoint from @xmath0 .
similarly , @xmath154 and @xmath158 cobound and annulus @xmath176 properly embedded in @xmath105 and disjoint from @xmath0 . since @xmath177 and @xmath178 are both untangles then @xmath175 and @xmath176 are boundary compressible in @xmath179 and @xmath180 .
let @xmath181 be the disk in @xmath105 gotten by boundary compressing @xmath175 and @xmath182 be the disk in @xmath105 gotten by boundary compressing @xmath176 . both @xmath183 and @xmath184
are disjoint from @xmath185 . by the above claim , @xmath184 and
@xmath183 are essential in @xmath21 . since @xmath186 , @xmath187 , and both @xmath183 and @xmath184 are disjoint from @xmath185 , @xmath188 .
we summarize the results of the previous lemmas using the following definition .
[ admissdef ] a singular foliation @xmath48 for a closed surface @xmath19 with @xmath33-marked points is * admissible * if it is induced by the standard height function on @xmath6 via some morse embedding of @xmath19 into @xmath6 such that the following hold : 1)there do not exist standard saddles @xmath23 , @xmath24 , and @xmath25 in @xmath48 such that @xmath23 is adjacent to @xmath24 and @xmath24 is adjacent to @xmath25 .
2)every outermost disk of @xmath39 contains at least one marked point .
* proof of theorem [ thm : struc ] * by lemma [ iness ] and lemma [ triple ] , @xmath48 is admissible when we view @xmath19 as a surface with marked points @xmath45 . for a fixed surface type and number of puncture points ,
the number of saddles in any admissible singular foliation is bounded .
[ numberofsaddles ] if @xmath19 is topologically a 2-sphere with @xmath33-marked points and @xmath48 is admissible , then the number of saddles in @xmath48 is at most @xmath189 .
let @xmath48 be an admissible singular foliation for @xmath19 that has a maximal the number of saddles .
claim : all marked points are contained in the collection of outermost disks of @xmath48 .
proof of claim : suppose not .
hence , there is a puncture point @xmath190 not on an outermost disk of @xmath48 .
there is a non - singular leaf @xmath59 in @xmath48 containing @xmath190 .
isotope @xmath19 in a neighborhood of @xmath59 as in figure [ fig : admiss1.eps ] .
the resulting singular foliation remains admissible but has one more saddle then @xmath48 , contradicting the maximality of the number of saddles of @xmath48 .
@xmath173 claim : all outermost saddles of @xmath48 contain exactly one puncture point .
proof of claim : suppose not .
hence , there is an outermost disk @xmath81 that contains two puncture points @xmath190 and @xmath191 .
let @xmath192 and @xmath193 be the closed curves in @xmath48 containing @xmath190 and @xmath191 respectively . up to relabeling
we can assume that @xmath193 bounds a monotone disk in @xmath48 that contains @xmath190 .
alter @xmath48 in a neighborhood of @xmath193 as in figure [ fig : admiss2.eps ] .
the resulting singular foliation remains admissible but has one more saddle then @xmath48 , contradicting the maximality of the number of saddles of @xmath48 .
@xmath173 we proceed by induction on @xmath194 .
suppose @xmath195 , then the maximal number of saddles is two , as depicted in figure [ fig : admiss3.eps ] .
assume @xmath39 has at most @xmath196 saddles for @xmath197 .
we will show @xmath39 has at most @xmath198 saddles for @xmath199 .
choose @xmath48 to have the maximal number of saddles of any admissible singular foliation .
since all puncture points are contained in outermost disks of @xmath48 and all outermost disks of @xmath48 contain exactly one puncture point , then @xmath39 contains exactly @xmath200 non - standard saddles . since we can assume @xmath201 there exists a non - standard saddle for @xmath39 .
label this saddle @xmath23 . by passing to an outermost non - standard saddle and possibly relabeling @xmath153 , @xmath154 and @xmath202
we can assume that both @xmath153 and @xmath154 bound disks @xmath203 and @xmath204 respectively in @xmath48 such that both @xmath203 and @xmath204 are disjoint from all non - standard saddles .
each of @xmath203 and @xmath204 meets a unique outermost disk of @xmath48 . by condition @xmath205 of the definition of admissible and the previous claim ,
each of these disks contains a unique marked point . by condition @xmath206 of the definition of
admissible , each of @xmath203 and @xmath204 contains exactly two saddles , all of which are standard .
replace a neighborhood of the union of @xmath203 and @xmath204 in @xmath48 with a single monotone disk @xmath105 containing a unique marked point .
call the resulting singular foliation @xmath207 and notice that @xmath207 is a foliation for a sphere with @xmath208 marked points .
notice that @xmath207 has 5 fewer saddles than @xmath48 .
if we can show that @xmath207 is admissible , then , by the induction hypothesis , @xmath48 has at most @xmath209 saddles and we have proven the theorem .
beginning with an embedding of a sphere with @xmath210 marked points in @xmath6 realizing @xmath211 , use the isotopy from lemma [ iness ] to eliminate the outermost saddles in @xmath203 and @xmath204 and iterate this process until @xmath23 is outermost and can be eliminated similarly .
since the isotopy in lemma [ iness ] preserves the singular foliation induced by the height function outside a neighborhood of @xmath212 , then this process produces an embedding of a sphere with @xmath208 marked points into @xmath6 with induced singular foliation @xmath207 .
if @xmath207 fails definition [ admissdef ] criteria @xmath206 , then , by inclusion of the complement of @xmath105 in @xmath207 into @xmath48 , @xmath48 also fails criteria @xmath206 , a contradiction .
let @xmath213 be an outermost disk of @xmath207 .
if @xmath213 is again an outermost disk in @xmath39 via inclusion , then @xmath213 contains one marked point by the admissibility of @xmath48 .
if @xmath213 is not an outermost disk for @xmath39 , then @xmath213 contains @xmath105 and , thus , contains a marked point .
hence , @xmath214 is admissible .
in this section we define tangle product and use the combinatorial result of the previous section to show that , under suitable hypothesis , the bridge number of a tangle product is supper - additive up to constant error . a graph @xmath215 is an @xmath7-star graph if @xmath215 has @xmath7 edges and @xmath216 vertices such that @xmath7 of the vertices are valence one and one of the vertices is valence @xmath7 .
denote by @xmath217 the set of valence one vertices .
let @xmath8 and @xmath9 be links embedded in distinct copies of @xmath6 , @xmath218 and @xmath219 .
let @xmath220 and @xmath221 be @xmath7-star graphs embedded in @xmath218 and @xmath219 respectively such that @xmath222 .
let @xmath223 be a small , closed , regular neighborhood of @xmath224 in @xmath225 such that @xmath226 is a rational tangle .
let @xmath227 .
a link in @xmath6 obtained by gluing @xmath228 to @xmath229 via a homeomorphism such that points in @xmath230 are mapped to points in @xmath231 is called an * @xmath7-strand tangle product * of @xmath8 and @xmath9 and is denoted by @xmath18 .
the image of @xmath228 and @xmath229 under this identification is called the product sphere and is denoted @xmath19 .
let @xmath21 be a minimal bridge sphere for @xmath18 of distance at least three .
we can assume that @xmath21 is @xmath22-level and that @xmath19 is taut with respect to @xmath21 . by theorem [ thm : struc ] ,
@xmath48 is admissible . by lemma [ numberofsaddles ]
, @xmath48 contains at most @xmath232 saddles .
if the set of saddles @xmath48 is nonempty , then it contains at least two outermost disks .
hence , there exists an outermost saddle @xmath23 such that @xmath81 meets @xmath18 in at most @xmath7 points .
we can eliminate @xmath23 via the ambient isotopy that horizontally shrinks and vertically lowers @xmath85 as in the proof of lemma [ iness ] .
this isotopy produces at most @xmath7 new maxima for @xmath233 .
see figure [ fig : createmax.eps ] . by repeating this process
, we can eliminate all saddles of @xmath48 at the cost of creating at most @xmath7 new maxima per saddle .
thus , we can assume we have an embedding of @xmath18 with at most @xmath234 maxima such that @xmath48 contains no saddles .
denote this embedding of @xmath18 by @xmath121 . if @xmath48 contains no saddles ,
there is a level preserving isotopy of @xmath84 taking @xmath19 to a standard round 2-sphere .
such an isotopy preserves the number of maxima of @xmath235 . recall that @xmath19 decomposes @xmath6 into two 3-balls @xmath236 and @xmath237 .
the link @xmath1 can be recovered from the tangle @xmath238 by glueing a trivial tangle @xmath239 to @xmath238 along their common @xmath10-punctured sphere boundary .
see figure [ fig : addingrational2.eps ] .
the number of maxima of the resulting embedding of @xmath1 is at most @xmath7 more than the number of maxima of @xmath240 in @xmath123 . by a similar argument
, we can produce an embedding of @xmath2 with at most @xmath7 more maxima than the number of maxima of @xmath240 in @xmath124 .
hence , @xmath241 , or @xmath20 . with more detailed analysis the constant @xmath242 that appears in the statement of theorem [ main ] can be improved
. however , the author believes it can not be improved beyond a quadratic expression in @xmath7 using the techniques presented in this paper . | we show that essential punctured spheres in the complement of links with distance three bridge spheres have bounded complexity .
we define the operation of tangle product , a generalization of both connected sum and conway product .
finally , we use the bounded complexity of essential punctured spheres to show that the bridge number of a tangle product is at least the sum of the bridge numbers of the two factor links up to a constant error .
[ section ] [ thm]lemma [ thm]definition [ thm]remark [ thm]conjecture [ thm]corollary [ thm]main result [ thm]assumption [ thm]proposition [ thm]question |
since the time when wheeler coined the name _ superheavy nuclei _ ( shn ) in 1955 @xcite , our knowledge of the very heavy nuclei " has become more extensive and systematic . during the last 60 years
the heaviest known nucleus limit has been extended from @xmath3md ( 1955 ) @xcite to @xmath4og ( 2006 ) @xcite , and the properties of shn have been studied mostly in the region of prolate deformations . in this region ,
the energy surfaces of shn reveal two paths to fission : a reflection - symmetric path corresponding to elongated fission fragments ( sef ) and the reflection - asymmetric path with elongated fission fragments ( aef ) , which bifurcates from the sef path after the first barrier , see e.g. ref .
there are also predictions on the ground state deformations , fission barrier heights , and spontaneous - fission and @xmath5-decay half - lives of shn ; for a recent review , see for example refs .
@xcite .
theoretically , the properties of shn have been studied much less in the oblate region than in the prolate region , with a few exceptions , such as the study on super - deformed - oblate shn at quadrupole moment @xmath6 to @xmath7 b @xcite . within the self - consistent constraint skyrme - hartree - fock+bcs model
, we found equilibrium toroidal nuclear density distributions at oblate deformation @xmath8 b for the hypothetical shn @xmath9 , @xmath10 , @xmath11 , and @xmath12 @xcite .
it is interesting to note that it was also wheeler who suggested long ago that under appropriate conditions the nuclear fluid may assume a toroidal shape @xcite . in 1970
s the idea of a toroidal nucleus was examined in the framework of the liquid drop model and shell corrections @xcite .
this contribution is devoted to a systematic investigation on the chain of even - even @xmath0 isotopes and @xmath1 isotones within the self - consistent constraint skyrme - hartree - fock - bogoliubov ( skyrme - hfb ) mean - field theory at the region of large oblate deformations .
the constrained skyrme - hfb approach is equivalent to the minimization of the skyrme energy density functional @xmath13 $ ] with respect to the densities and currents under appropriate constraints @xcite . using the method of lagrange multipliers we solve an equality - constrained problem ( ecp ) : @xmath14\\ \mbox{subject to : } \displaystyle\langle \hat{n}_{q } \rangle= n_{q},\quad ( q = p
, n),\\ \phantom{\mbox{subject to : } } \displaystyle\langle \hat{q}_{\lambda\mu } \rangle= q_{\lambda\mu } , \end{array } \right .
% \nonumber \label{eq:1}\ ] ] where the constraints are defined by the average values @xmath15 of the proton and neutron particle - number operators @xmath16 and by the constrained values @xmath17 of the mass - multiple - moment operators @xmath18 .
the above ecp equations were solved using an augmented lagrangian method @xcite with the symmetry - unrestricted code hfodd @xcite . in the particle - hole channel
the skyrme skm * @xcite force was applied and a density - dependent mixed pairing interaction @xcite in the particle - particle channel was used .
the code hfodd uses the basis expansion method utilizing a three - dimensional cartesian deformed harmonic oscillator basis . in the present study
, we used a basis which consists of states having not more than @xmath19 quanta in the cartesian directions , and not more than 1140 states .
as an example , the total hfb energy of sh nucleus @xmath20 as a function of the quadrupole moment @xmath2 is shown in fig .
[ fig:1 ] .
in addition to a spherical ground state minimum , one can see two paths leading to fission on the prolate side : a reflection - symmetric path with the elongated fission fragments ( sef ) ( open circles ) and a reflection - asymmetric path with the elongated fission fragments ( aef ) ( dashed line ) . on the oblate side , the self - consistent nuclear density under the @xmath2 constraint changes from an oblate spheroidal to a biconcave disc shape , as the magnitude of oblate @xmath2 increases .
when the oblate @xmath2 magnitude exceeds 158 b , there emerges an additional self - consistent toroidal nuclear density solution .
( color online . )
total hfb energy of @xmath20 as a function of the quadrupole moment . the open circle symbols and
dashed ( blue ) line show the symmetric ( sef ) and asymmetric ( aef ) elongated fission pathways , respectively .
the axially symmetric sef ( @xmath21 ) fission pathway is marked by a solid thin line .
the nuclear matter density distributions with toroidal shapes appear in the region of large oblate deformation @xmath22 b as ( red ) solid circles .
, scaledwidth=70.0% ] an enlarged view of the transition from the biconcave disc to the toroidal shape of fig .
[ fig:1 ] is shown in fig .
[ fig:2ab](a ) , where our skyrme - hfb calculations give the biconcave disc solutions ending at the point * a * ( at @xmath23 b ) , and another toroidal solutions starting at the point * b * ( at @xmath24 b ) .
the nuclear density distributions of @xmath20 calculated at * a * and * b * are depicted in fig .
[ fig:2ab](b ) , which indicates that the nuclear density at the toroidal geometrical center is very small at the starting point * b * of the toroidal sequence , but remains to be about 0.08/@xmath25 at point * a * of the biconcave disc sequence .
( color online . )
left panel ( a ) , an enlarged view of fig .
[ fig:1 ] , the total hfb energy of @xmath20 as a function of the quadrupole moment between points * a*-*b * , where a shape transition from biconcave discs to toroidal shapes takes place .
right panel ( b ) , the nuclear density distribution at @xmath26 b ( biconcave disc ) and @xmath24 b ( torus ) . ]
( color online . )
left panel ( a ) , an enlarged view of fig .
[ fig:1 ] , the total hfb energy of @xmath20 as a function of the quadrupole moment between points * a*-*b * , where a shape transition from biconcave discs to toroidal shapes takes place .
right panel ( b ) , the nuclear density distribution at @xmath26 b ( biconcave disc ) and @xmath24 b ( torus ) . ] in the region of quadrupole moment between @xmath27 and @xmath28 , both biconcave disc and toroidal solutions coexist for @xmath20 .
it is of interest to examine the single - particle states of these two types of solutions in this oblate deformation region .
the proton single - particle levels of @xmath20 close to the fermi energy as a function of @xmath2 between @xmath27 and @xmath28 for the biconcave disc and toroidal sequences are shown in the upper and lower panels of fig .
[ fig:3 ] , respectively .
levels with positive parity are drawn as solid curves , while those with negative parity are drawn as dashed curves .
each single - particle state is labeled by the nilsson quantum numbers @xmath29\omega$ ] of the dominant component .
each level is doubly degenerate .
the circled numbers denote the occupation numbers . for the sake of comparison , fig .
[ fig:4 ] gives the neutron single - particle levels close to the fermi energy for the biconcave disc solution in the upper panel and the toroidal solution in the lower panel .
-90 ( color online . )
proton single - particle levels of @xmath20 as a function of the quadrupole moment .
the levels with positive parity are drawn with solid lines , while those with negative parity are drawn with ( blue ) dashed lines .
the upper panel is for a biconcave disc shape and the lower panel for a toroidal shape .
, title="fig : " ] + -90 ( color online . ) proton single - particle levels of @xmath20 as a function of the quadrupole moment .
the levels with positive parity are drawn with solid lines , while those with negative parity are drawn with ( blue ) dashed lines .
the upper panel is for a biconcave disc shape and the lower panel for a toroidal shape .
, title="fig : " ] even though figs .
[ fig:3 ] and [ fig:4 ] pertain to the self - consistent single - particle states for @xmath30 , we expect that the mean - field potential depends mostly on nuclear density shape and the quadrupole moment , and varies only slightly as a function of the atomic number and the neutrons number .
the single - particle state diagrams in figs .
[ fig:3 ] and [ fig:4 ] can be approximately applied to an extended region around @xmath20 .
-90 ( color online . ) the same as in fig .
[ fig:3 ] , but for the neutron single - particle levels .
, title="fig : " ] + -90 ( color online . ) the same as in fig .
[ fig:3 ] , but for the neutron single - particle levels .
, title="fig : " ] we infer from figs . [ fig:3 ] and [ fig:4 ] that the densities of proton and neutron single - particles states are far from being uniform .
there are regions of sparse density of single - particle states which can be identified as single - particle shells " associated with enhanced stability @xcite .
it will be of future interest to exploit the property of the extra stability of shn for which the toroidal proton and neutron shells are located at the same deformation .
the shell effects play an even more significant role when the bulk properties of the system lead to a nearly flat bulk energy as a function of the deformation , such as would be expected for systems with @xmath31 @xcite .
the complexity of the single - particle state energy level diagram indicates that the location of the neutron and proton shells needs to be examined on a case - by - case basis .
( color online . )
shape transitions from the biconcave disk shapes to the toroidal shapes for the even - even isotopes @xmath0 in panel ( a ) , and for the even - even isotones @xmath1 in panel ( b ) .
the points of energy crossing of the oblate spheroids and toroidal shapes solutions are connected by a(blue ) dashed line . ]
( color online . )
shape transitions from the biconcave disk shapes to the toroidal shapes for the even - even isotopes @xmath0 in panel ( a ) , and for the even - even isotones @xmath1 in panel ( b ) .
the points of energy crossing of the oblate spheroids and toroidal shapes solutions are connected by a(blue ) dashed line . ]
[ fig:7ab](a ) gives the skyrme - hfb energies as a function of the quadrupole moment of even - even @xmath0 isotopes with the number of neutrons from 166 to 190 , in the region of the shape transition from the biconcave disc shape to the toroidal shape .
the toroidal and biconcave disc total energies decreases as a function of increasing @xmath2 and do not posses an energy minimum .
the slope of the toroidal energy as a function of @xmath2 appears to be nearly independent of the neutron number , for @xmath0 isotopes .
the biconcave disc energy curve and the toroidal energy curve cross each other at an energy crossing point , whose location moves to a more negative @xmath2 value as the neutron number increases . along the lowest energy curve as a function of @xmath2 in fig . [ fig:7ab](a ) , there is a sudden shape transition from the biconcave disc shape to the toroidal shape at the energy crossing point .
the first solution with the toroidal shape takes place at @xmath32 b for @xmath33 , and at @xmath34 b for @xmath35 .
similar results , but for the even - even @xmath1 isotones with the number of protons from 106 to 124 are shown in fig .
[ fig:7ab](b ) . for those isotones
, one observes that as the proton number increases , the magnitude of the coulomb repulsion increases , and the magnitude of the slope of the toroidal energy curve becomes smaller .
the toroidal energy curve for @xmath36 is nearly but not completely flat .
further increase in the proton number may render a toroidal energy equilibrium at a greater oblate deformation .
the energy crossing points of the biconcave disk and toroidal energy curves occur at @xmath37 b for all atomic numbers in fig .
[ fig:7ab](b ) .
the coulomb repulsion from a large number of protons in a sh nucleus has a tendency to push the nuclear matter outward , making it easier to assume a toroidal shape .
we examine here shn in the region of @xmath0 isotopes and @xmath1 isotones .
we find that as the magnitude of the oblate @xmath2 increases along the lowest energy curve , there is a sudden shape transition from a biconcave disc to a torus .
for @xmath0 isotopes with @xmath38 and for @xmath1 isotones with @xmath39 , the total energy curves lie on a slope , indicating that these nuclei in a toroidal shape are unstable against returning to the shape of a sphere - like geometry . our examination of the single - particle states in this region
reveal that the density of single - particle levels is far from being uniform and single - particle shells are present at various toroidal deformations . because the energy curve as a function @xmath2 becomes flatter with increasing @xmath40 , one expects that by increasing the atomic number beyond @xmath31 with possible appropriate toroidal shells , some toroidal figures of equilibrium may become possible .
future search for toroidal nuclei may focus attention in this superheavy region in conjunction with possible nuclear shell effects . the presence of a large angular momentum will facilitate the formation of a toroidal nucleus . in this regard
, it will be useful to examine shn with non - collective rotations whose spin along the symmetry axis arises from particle - hole excitations @xcite .
previous investigation reveal a region of toroidal high - spin isomers in the light mass region @xcite .
a recent investigation shows that @xmath20 with @xmath41 and 208@xmath42 may be toroidal high - spin isomers @xcite .
the research was supported in part by the division of nuclear physics , u.s .
department of energy under contract de - ac05 - 00or22725 and the national science center , poland , project no 2016/21/b / st2/01227 . | using the hartree - fock - bogoliubov ( hfb ) self - consistent mean - field theory with the skm * skyrme energy - density functional , we study nuclear structure properties of even - even superheavy nuclei ( shn ) of @xmath0 isotopes and @xmath1 isotones . the shape of the nucleus along the lowest energy curve as a function of the quadrupole moment @xmath2 makes a sudden transition from the oblate spheroids ( biconcave discs ) to the toroidal shapes , in the region of large oblate quadrupole moments . |
we start with a regular lattice ; on each site @xmath4 lives a random energy @xmath5 , drawn independently from site to site from a common distribution @xmath6 .
we assume @xmath6 is continuous and normalized to unity , i.e. @xmath7 . for any choice of the set of @xmath5 s we obtain an energy landscape , i.e. , a topological space ( defined through nearest neighbors on the lattice ) with an energy for each element . for a given realization of the landscape ,
a site @xmath4 is a local minimum if @xmath8 where @xmath9 denotes any nearest neighbour sites of @xmath4 .
hereafter we shall denote by @xmath2 the number of local minima on this landscape .
for any given realization of the landscape , one can formally express @xmath2 as @xmath10 where the product runs over all nearest neighbours @xmath9 of site @xmath4 and @xmath11 is the heaviside theta function .
we are interested in computing the probability distribution @xmath12 of the total number of local minima ( or equivalently that of the maxima ) .
the first important observation is that the distribution @xmath12 is completely independent of the energy distribution @xmath6 . to see this
, we formally express the distribution as @xmath13 where @xmath14 is the dirac delta function .
next we make a change of variable for each @xmath15 @xmath16 clearly @xmath17 is a monotonically increasing function of @xmath18 .
therefore , @xmath19 . moreover , since @xmath6 is normalized to unity , the variable @xmath20 varies from @xmath21 to @xmath22 and so eq . ( [ m2 ] ) simply becomes @xmath23 we see that the energy distribution @xmath6 simply drops out and @xmath12 , for arbitrary @xmath6 , is universal and is the same as when the ` new ' energy variable @xmath24 is drawn independently from a uniform distribution over @xmath25 $ ] .
thus , the model is simplified : at each lattice site lives a random number @xmath26 $ ] drawn independently from a uniform distribution
. we will refer to this model as the random minima model , for which we want to compute the distribution @xmath12 of the total number of local minima .
following the mapping to the uniform distribution , it follows that the random minima model is just the continuous version of the permutation generated landscape recently studied by hivert et .
they considered a set of integers @xmath27 .
each of the @xmath28 permutations of this set defines a random energy landscape and they occur with equal probability . thus the energy at a site is now a discrete integer drawn uniformly from the set @xmath29 $ ] with equal probability @xmath30 .
thus the statistics of the minima in the permutation generated landscape will be identical to that of the random minima model .
the mean and the variance of @xmath2 are relatively straightforward to compute on an arbitrary lattice since they depend only on the local properties of the landscape . to see this ,
let us define the variable @xmath31 where @xmath9 runs over the nearest neighbours of @xmath4 and the @xmath24 s are independent random numbers in @xmath32 $ ] drawn from the uniform distribution .
this @xmath33 is an indicator function which is 1 if site @xmath4 is a local minimum and 0 otherwise .
then it follows that @xmath34 taking the average in eq .
( [ mean1 ] ) and using the translational invariance of the lattice , it follows that @xmath35^n = \frac{1}{n+1 } \label{mean2}\ ] ] where @xmath36 is the number of nearest neighbours of any site , i.e. , the co - ordination number of the lattice .
this result also follows trivially from a combinatorial argument on the permutation landscape : consider the site @xmath4 with its @xmath36 neighbours .
the number of ways one can arrange the integers on these @xmath37 sites with the restriction that @xmath4 is a local minimum is @xmath38 clearly . on the other hand , the total number of unrestricted configurations is @xmath39 so the probability that the site @xmath4 is a minimum is simply @xmath40 .
the calculation of the variance takes a few more steps . squaring eq .
( [ mean1 ] ) and taking the average gives @xmath41 note that the @xmath33 s are correlated random variables but only over a short range .
the calculation of the correlation function @xmath42 can therefore be performed by hand ( it only involves the calculation of simple integrals involving a maximum of @xmath43 sites ) .
for example , in @xmath22-d , one gets @xmath44 this result has been obtained in various contexts before , such as in the calculation of the number of metastable states in @xmath22-d ising spin glass @xcite and also in the context of the occupation time of a non - markovian sequence in @xmath22-d @xcite .
recently , the variance for the @xmath45 square lattice was computed in the permutation model by hivert et .
al . @xcite . for large @xmath3
, one gets @xmath46 in general , on any arbitrary lattice , for large @xmath3 @xmath47 where @xmath48 and @xmath49 are lattice dependent numbers that can be computed either using integrals or by combinatorics in the permutation model . near the mean @xmath50 and within a region @xmath51 , the distribution @xmath12 is expected to be a gaussian @xcite with mean and variance given in eq .
( [ mv ] ) @xmath52 . \label{gauss1}\ ] ] however , for @xmath2 far from the mean , one would expect deviations in @xmath12 from the gaussian form . in this paper
, we are interested in the probabilities of such large deviations , i.e. , we are interested in computing the probabilities of occurrences of configurations that are far from typical . on general grounds , in the limit @xmath53 , @xmath54 but with their ratio @xmath55 fixed but arbitrary
, one expects that @xmath12 has the form @xmath56 \label{ld1}\ ] ] where @xmath57 is a large deviation function .
now , on a given lattice , @xmath2 can take values from @xmath21 to a maximal number @xmath58 .
the upper limit follows from the fact that once a site is a local minimum , none of its nearest neighbours can be a local minimum .
thus there is a nearest neighbour exclusion principle for the minima .
this constraint indicates that one can not pack arbitrarily large number of minima on the lattice and there is an upper bound on the number of minima .
for example , on a bipartite lattice , one can pack at most @xmath59 local minima , one at every alternate site .
in particular , on a square lattice the minima will be placed on a checkerboard pattern .
thus , in this case , the large deviation function @xmath57 is defined for @xmath60 . on the other hand , on a cayley tree with @xmath61 number of branches and @xmath3 sites
, we will see later that @xmath62 , thus @xmath63 .
as mentioned above , in the vicinity of its mean , i.e. , for @xmath64 , the distribution @xmath12 is gaussian as in eq .
( [ gauss1 ] ) .
this indicates that the large deviation function @xmath57 is quadratic near @xmath65 @xmath66 such that @xmath67 \sim \exp\left[-(m - an)^2/{2bn}\right]$ ] has the required gaussian form .
however , far from the mean , @xmath12 will have non - gaussian tails , indicating a departure of @xmath57 from the simple quadratic form .
a knowledge of the function @xmath57 would then allow one to compute the probabilities of occurrences of atypical configurations , such as the probability of a configuration with the lowest number of minima ( e.g. @xmath68 ) or the ones with the maximal number of minima ( @xmath69 .
a particular focus of this paper will be to compute the probability of a maximally packed configuration , i.e. , when @xmath70 .
it follows from the general form in eq .
( [ ld1 ] ) that the probability of this maximal packing should decay exponentially with the system size for large @xmath3 , @xmath71 \sim \gamma^{-n } \label{ld2}\ ] ] where @xmath72 $ ] is a lattice dependent constant .
of course , this expression also gives the scaling of the fraction of permutations which have @xmath73 local minima in the permutation landscape .
it is then natural to interpret @xmath74 as the entropy cost per site to achieve maximally packed minima .
we shall compute the constant @xmath75 exactly for a number of lattices . for the @xmath76 chain
, we will show that @xmath77 note that this result appeared before in the context of metastable states in @xmath76 ising spin glasses at zero temperature @xcite .
furthermore , the same @xmath78 decay also appeared as the persistence probability of a @xmath76 non - markovian sequence @xcite . for the @xmath76 case
, one can also compute the full large deviation function @xmath57 ( see section [ sec : onedimension ] ) . another lattice where we can calculate @xmath75
exactly is the cayley tree with @xmath61 branches where we show that @xmath79 where @xmath80 is the beta function .
as expected , for @xmath81 , it reduces to the @xmath22-d result in eq .
( [ g1d ] ) . as @xmath61 increases
, @xmath75 increases slowly and as @xmath82 , @xmath83 .
a cayley tree with an infinite number of branches corresponds to a hypercubic lattice in @xmath0-dimensions in the @xmath84 limit .
thus , based on these two limiting results , one expects the constant @xmath75 on any @xmath0-dimensional lattice to satisfy the bounds @xmath85 our numerical simulations for @xmath86 are consistent with these bounds .
we have also been able to compute @xmath75 exactly for a two - leg ladder where we show that @xmath87^{1/3}=1.57657\dots \label{glad}\ ] ] where @xmath88 is the smallest positive root of the bessel function @xmath89 .
the result in eq .
( [ glad ] ) is , of course , consistent with the general bounds in eq .
( [ bounds ] ) .
we now consider these successive cases in detail .
in a @xmath22-d chain of size @xmath3 , the distribution @xmath12 of the total number of local minima was computed exactly by derrida and gardner @xcite , although they did not calculate the large deviation function @xmath57 explicitly . however , from their result for the generating function , it is easy to derive the large deviation function by a legendre transform . for the sake of completeness , we provide here a brief derivation of the @xmath22-d result , albeit by a slightly different method .
it is useful to define the generating function or the partition function @xmath90 where @xmath91 is the fugacity or the weight associated with each minimum .
for simplicity we consider an open chain of size @xmath3 and let @xmath92 be the value of the random variable at the @xmath3-th site . to write a recursion relation for the partition function , it is convenient to define two restricted partition functions : @xmath93 and @xmath94 denoting respectively the partition functions conditioned on the fact that the @xmath3-th site has value @xmath92 and that it is respectively a local minimum ( i.e. , @xmath95 ) or a local maximum ( @xmath96 ) .
note that since we are considering an open chain , the last site ( @xmath3 ) has only one neighbour to its left , namely the @xmath97-th site .
knowing the restricted partition functions , one can compute the full partition function from the relation @xmath98 .
\label{pf1d}\ ] ] the restricted partition functions satisfy a pair of simple recursion relations @xmath99\ , dy .
\label{recur2}\end{aligned}\ ] ] these recursion relations can be easily understood by considering all possibilities when one adds a new site to the chain .
if the new site is a minimum , we need to attach a factor @xmath91 . on the other hand ,
if the @xmath97-th site was a minimum and it ceases to be a minimum after the addition of the @xmath3-th site , we have to detach a factor @xmath91 .
it follows from eqs .
( [ recur1 ] ) and ( [ recur2 ] ) that the restricted partition functions satisfy the boundary conditions : @xmath100 and @xmath101 . for large @xmath3
, one expects a separation of variables between @xmath20 and @xmath3 of the form , @xmath102 where @xmath103 is a function of @xmath91 only ( but independent of @xmath20 ) and the functions @xmath104 and @xmath105 satisfy the boundary conditions : @xmath106 and @xmath107 . substituting this ansatz in eqs .
( [ recur1 ] ) and ( [ recur2 ] ) and subsequently differentiating with respect to @xmath20 , we get a pair of differential equations @xmath108 diagonalizing the @xmath109 $ ] matrix , one obtains the solutions @xmath110 where @xmath48 and @xmath49 are arbitrary constants .
the two boundary conditions @xmath106 and @xmath107 yield two relations between @xmath48 and @xmath49 @xmath111 eliminating @xmath48 and @xmath49 between eqs .
( [ rel1 ] ) and ( [ rel2 ] ) determines the eigenvalue @xmath103 exactly @xmath112,\quad\quad { \rm for}\quad\quad 0\le z \le 1 \label{eigen1 } \\ & = & \frac{1}{\sqrt{z-1 } } \tan^{-1 } \left[\sqrt{z-1}\right ] \quad\quad { \rm for}\quad\quad z \ge 1 .\label{eigen2}\end{aligned}\ ] ] note that the function @xmath103 is analytic at @xmath113 .
the form in eq .
( [ eigen2 ] ) is just an analytical continuation of the form in eq .
( [ eigen1 ] ) for @xmath114 . substituting the large @xmath3 form of @xmath115 and @xmath116 in eq .
( [ pf1d ] ) , one obtains the large @xmath3 behavior of the partition function , @xmath117^{-n } \label{pfn1}\ ] ] where @xmath103 is given in eqs .
( [ eigen1 ] ) and ( [ eigen2 ] ) . substituting the anticipated form of @xmath118 $ ] in eq .
( [ pfn1 ] ) one gets @xmath119\sim [ \lambda(z)]^{-n}. \label{pfn2}\ ] ] taking the @xmath54 limit in eq .
( [ pfn2 ] ) gives @xmath120 = \ln ( \lambda(z ) ) .
\label{pfn3}\ ] ] inverting via the legendre transform finally gives the large deviation function @xmath121 , \label{ld3}\ ] ] where @xmath103 is given in eqs .
( [ eigen1 ] ) and ( [ eigen2 ] ) .
note that determining @xmath57 from eq .
( [ ld3 ] ) requires a knowledge of @xmath103 for all @xmath122 , thus we need both formulae of @xmath103 in eqs .
( [ eigen1 ] ) and ( [ eigen2 ] ) . for @xmath123 in @xmath22-d obtained from eq .
( [ ld3 ] ) using mathematica.,width=377,height=264 ] we have obtained @xmath57 from eq .
( [ ld3 ] ) using mathematica and it is displayed in fig .
[ fig : ldf1d ] .
since the maximal value of @xmath2 in @xmath22-d is @xmath124 , the allowed range of @xmath125 is @xmath123 .
one can analytically obtain the form of @xmath57 in the three limiting cases @xmath126 , @xmath127 and @xmath128 .
first let us consider the limit @xmath126 . to find @xmath57 in this limit we need to use the @xmath129 form of @xmath103 in eq .
( [ ld3 ] ) . as @xmath129 , it is easy to see from eq .
( [ eigen1 ] ) that @xmath130 . substituting this behavior in eq .
( [ ld3 ] ) and subsequently maximizing the r.h.s of eq .
( [ ld3 ] ) we find that @xmath57 diverges logarithmically , @xmath131 $ ] as @xmath126 .
next we consider the @xmath127 limit .
note that the mean number of minima @xmath132 in @xmath133 which follows from the general result in eq .
( [ mean2 ] ) .
thus @xmath127 limit corresponds to behavior of @xmath2 near its mean and one expects a quadratic form for @xmath57 near @xmath134 .
indeed this also follows from eq .
( [ ld3 ] ) .
the limit @xmath127 corresponds to using the @xmath135 behavior of @xmath103 in eq .
( [ ld3 ] ) .
substituting @xmath136 in eq .
( [ eigen2 ] ) and expanding in powers of @xmath137 , one gets @xmath138 . substituting this result on the r.h.s of eq .
( [ ld3 ] ) and maximizing one gets the expected quadratic behavior , @xmath139 near @xmath127 .
this is thus a special case in @xmath22-d of the general behavior in eq .
( [ qua ] ) with @xmath140 and @xmath141 .
finally , to derive the maximally packed limit @xmath128 , we need to use the @xmath142 behavior of @xmath103 in eq .
( [ ld3 ] ) . in this limit
, it follows from eq .
( [ eigen2 ] ) that @xmath143 .
using this form and maximizing the r.h.s .
( [ ld3 ] ) we get @xmath144 as @xmath128 .
thus , summarizing the three limiting behaviors @xmath145 \quad\quad { \rm as}\quad\quad y\to 0 \nonumber \\ & \approx & \frac{45}{4}\ , ( y-1/3)^2 \quad\quad { \rm as}\quad\quad y\to 1/3 \nonumber \\ & \approx & \ln ( \pi/2 ) + 2(1/2-y)\ , \ln(\pi(1/2-y)/e ) \quad\quad { \rm as}\quad\quad y\to 1/2 .
\label{limiting}\end{aligned}\ ] ] note that as one approaches the maximally packed limit @xmath128 , @xmath146 .
thus , it follows from eq .
( [ ld2 ] ) that in @xmath22-d , for large @xmath3 @xmath147 the result declared in eq .
( [ g1d ] ) .
in this section , we focus only on the maximally packed configuration ( where the number of minima on the lattice is maximal ) . from the general large deviation theory ,
we have already argued that the probability of such a configuration @xmath148 is expected to decay exponentially with the system size @xmath3 as in eq .
( [ ld2 ] ) .
the goal is to compute the nontrivial constant @xmath75 . in the previous section
, we have shown that in a @xmath22-d chain , @xmath149 . in this section ,
we compute @xmath75 exactly in a few other solvable cases , notably for a cayley tree with @xmath61 branches and also for a two - leg ladder in @xmath150-d .
we consider a cayley tree with @xmath61 branches and @xmath36 generations .
we label the generations by @xmath151 starting from the leaf sites at the bottom ( see fig . [
fig : tree ] ) .
the total number of sites on the tree is @xmath152 note that in the limit @xmath153 , the tree reduces to a @xmath22-d chain with @xmath154 sites .
branches and @xmath155 generations .
the layers are labelled @xmath156 starting with the bottom layer.,width=453 ] at each site of this tree resides a random number @xmath24 drawn independently from the uniform distribution over @xmath32 $ ] .
we want to calculate the probability of the configuration with the maximum possible number of local minima .
note that if a site is a local minimum , its neighbours can not be local minima . to find the maximally packed configuration , we note that the number of sites at the bottom layer ( @xmath157 ) increases exponentially with @xmath36 as @xmath158 .
thus , to achieve maximal packing , it is necessary to fill up the bottom layer ( the leaf sites ) with local minima .
then the layer just above the bottom layer ( @xmath159 ) is devoid of local minima .
the next layer @xmath160 ) can again be packed with local minima .
thus , the maximally packed configuration is the one where alternate layers are fully packed with minima , starting with a fully packed layer at the bottom , as shown in fig .
[ fig : tree ] .
the total number of minima @xmath73 in this maximally packed configuration depends on whether the number of generations @xmath36 is even or odd @xmath161 for @xmath36 even , the root is not a minimum whereas for @xmath36 odd , the root is a minimum . in either case , for @xmath36 large , the total number of minima in the maximally packed configuration is proportional to the total number of sites @xmath3 , @xmath162 having identified the maximally packed configuration , we will next compute the probability of its occurrence and show that for large @xmath3 @xmath163 for simplicity , let us assume that @xmath36 is even .
one can perform an identical calculation when @xmath36 is odd . for @xmath36 even , all the even layers ( labelled by @xmath164 with @xmath165 ) are devoid of local minima whereas all the odd layers ( labelled by @xmath166 with @xmath167 ) are fully packed with local minima .
our aim is to write a recursion relation . for this
, it is convenient to define two probabilities @xmath168 and @xmath169 defined respectively for even and odd layers .
we define @xmath168 as the probability that a subtree of @xmath170 generations ( counted from the bottom of the tree ) has a random variable @xmath20 at its root and that @xmath20 is _ not _ a local minimum . on the other hand
, @xmath169 is the probability that a subtree of @xmath171 generations ( again counted from the bottom of the tree ) has a random variable @xmath20 at its root and that @xmath20 is a local minima .
it is then easy to see that they satisfy the recursion relations @xmath172^{\mu } \label{p1 } \\
q_{2m+1}(x ) & = & \left [ \int_x^1 p_{2m}(y ) \ , dy\right]^{\mu } \label{q1}\end{aligned}\ ] ] where @xmath165 .
the recursions start with the initial condition @xmath173 .
since @xmath36 is even , the root of the full tree with @xmath36 generations is not a local minima and hence the probability of the full tree is just @xmath174 given that the value at the root is @xmath20 .
the probability of the maximally packed configuration is then obtained by integrating over @xmath20 at the root @xmath175 thus , we need to solve eqs .
( [ p1 ] ) and ( [ q1 ] ) and then substitute the solution for @xmath174 in eq . ( [ pmp1 ] ) to calculate @xmath176 . to solve the nonlinear recursion relations , it is convenient to define @xmath177^{1/\mu}$ ] and @xmath178^{1/\mu}$ ] . then the recursions in eqs .
( [ p1 ] ) and ( [ q1 ] ) become @xmath179 starting with @xmath180 for @xmath181 .
it follows from eqs .
( [ p2 ] ) and ( [ q2 ] ) that they satisfy the boundary conditions , @xmath182 and @xmath183 for all @xmath184 .
however , @xmath185 and @xmath186 are nonzero .
it is then useful to define the ratios , @xmath187 and @xmath188 so that @xmath189 , @xmath190 and @xmath191 , @xmath192 . in terms of the new functions ,
the recursions become @xmath193 starting with @xmath194 for @xmath181 . as @xmath195 increases
, we expect that the ratio functions @xmath196 and @xmath197 will approach their respective @xmath195-independent fixed point forms @xmath198 and @xmath199 .
this means that as @xmath200 , @xmath201 and @xmath202 and @xmath203 with the boundary conditions @xmath204 , @xmath205 and @xmath206 , @xmath207 . these boundary conditions will determine the eigenvalues @xmath208 and @xmath209 .
differentiating eqs .
( [ f2 ] ) and ( [ g2 ] ) with respect to @xmath20 gives @xmath210 multiplying ( [ f3 ] ) by @xmath211 and ( [ g3 ] ) by @xmath212 , adding and then integrating , we find a conserved quantity @xmath213 where @xmath214 is a constant independent of @xmath20 .
putting @xmath215 and using the respective boundary conditions gives @xmath216 .
thus , @xmath217 and @xmath218 the common eigenvalue @xmath219 is yet to be determined . eliminating @xmath199 between eqs .
( [ f3 ] ) and ( [ com2 ] ) gives @xmath220^{\mu/(\mu+1 ) } \label{f4}\ ] ] subject to the boundary conditions @xmath204 and @xmath205 .
integrating and using @xmath204 we get @xmath221 using the other boundary condition @xmath205 determines @xmath219 explicitly @xmath222 where @xmath223 is the standard beta function .
the eigenfunction @xmath198 in eq .
( [ f5 ] ) can be expressed as the solution of the equation @xmath224 = \frac{x}{\mu+1}\ , b\left(\frac{1}{\mu+1},\frac{1}{\mu+1}\right ) \label{f6}\ ] ] where @xmath225 $ ] is the hypergeometric function .
the other eigenfunction @xmath199 follows from eq .
( [ com2 ] ) , or simply from the symmetry @xmath226 . to check that the solutions to the original recursion relations ( [ p2 ] ) and ( [ q2 ] ) indeed converge to these fixed solutions , we have numerically solved eqs .
( [ p2 ] ) and ( [ q2 ] ) for @xmath227 .
we find that the numerical solutions ( the ratio functions @xmath196 and @xmath197 ) converge to fixed point functions rather quickly after about @xmath228 or @xmath229 iterations . in fig .
[ f3 ] , we compare the numerical fixed point solution ( the solution after @xmath230 iterations ) @xmath198 with the analytical solution in eq .
( [ f6 ] ) with @xmath227 .
the agreement is perfect .
after @xmath230 iterations , for @xmath227 , is compared to the analytical scaling function in eq .
( [ f6 ] ) obtained using mathematica .
the two curves are difficult to distinguish indicating perfect agreement.,width=377,height=264 ] to determine @xmath75 , we use eq .
( [ pmp1 ] ) and the recursion relation ( [ q2 ] ) which show that @xmath231 then we take ratios to extract the bulk contribution of the cayley tree , thereby removing artifacts coming from its surface @xcite : as @xmath232 , we have @xmath233 and @xmath234 . eliminating @xmath185 gives a recursion for large @xmath195 @xmath235 iterating it , we obtain , for large @xmath195 , @xmath236 . substituting this result in eq .
( [ pmp2 ] ) and using @xmath237 gives our final result for large @xmath3 @xmath238 note that the recursion relation ( [ q1 ] ) is valid only for large @xmath195 . but in deriving the results above we have assumed that it holds even for small @xmath195 . even though the asymptotic result is not expected to change due to this ` initial condition ' effect , one can avoid this `` surface effect '' by appropriately defining @xmath75 as the ratio @xmath239^{\mu } dx } { \left[\int_0 ^ 1 p_n(x)dx\right]^{\mu}}. \label{gnf1}\ ] ] this definition follows from the following observation .
the numerator on the r.h.s of ( [ gnf1 ] ) denotes the probability of a maximally packed configuration on a tree with @xmath37 generations .
the denominator is the joint probability that the disconnected @xmath61 subtrees of @xmath36 generations are all maximally packed .
thus , the ratio on the r.h.s . is just the factor by which the probability of a maximally packed configuration per site decreases when one fuses the @xmath61 number of @xmath36-generation trees and one additional root to construct a newly maximally packed configuration on a @xmath37-generation tree .
but , asymptotically for large @xmath36 , this is precisely @xmath240 by the orginal definition in eq .
( [ ld2 ] ) .
physically , @xmath74 is the additional entropy change when packing an extra minimum in the tree . using the recursion relations ( [ p1 ] ) , ( [ q1 ] ) in eq .
( [ gnf1 ] ) one gets @xmath241^{\mu } dx=\int_0 ^ 1 g_{n+1}^\mu(x ) dx . \label{gnf2}\ ] ] substituting @xmath242 , integrating and using @xmath205 , we get @xmath243 note that for @xmath81 , we recover the @xmath22-d result , @xmath149 . for @xmath227
, we get @xmath244 . as @xmath245
, @xmath75 converges slowly to @xmath246 .
since the @xmath245 result should coincide with that on a hypercubic lattice in the infinite dimension limit , we expect that for hypercubic lattices in @xmath0 dimensions , @xmath75 is bounded , @xmath247 .
as @xmath0 increases from @xmath22 to @xmath248 , @xmath75 should increase monotonically from its @xmath133 value @xmath249 to @xmath150 .
our numerical solutions on @xmath0-dimensional hypercubic lattices with @xmath250 are consistent with these bounds . on a bipartite lattice ,
the maximally packed configuration is the one where one places a local minimum at every alternate site .
for example , on a square lattice , a maximally packed configuration has a checkerboard pattern as shown in fig .
[ fig : plaque ] .
let us first focus on the neighbourhood of a single local minimum .
the sites at the corners of the plaquette containing this minimum are clearly not local minima .
let @xmath251 , @xmath252 , @xmath253 and @xmath254 denote the values of the random variables at the four corners of the plaquette .
then , given these four values , the probability that the site at the center of the plaquette is a local minimum is clearly @xmath255 where we have used the fact that the random variable at the center of the plaquette is drawn from a uniform distribution .
so , the probability of the full checkerboard configuration where every alternate site is a local minimum , given the values of the random variables @xmath256 s at the sites of the other sublattice , is obtained by multiplying all the plaquettes @xmath257 where the plaquettes are labelled by @xmath9 and @xmath258 , @xmath259 , @xmath260 and @xmath261 are the four random variables at the corners of the @xmath9-th plaquette .
the total number of plaquettes will also be denoted by @xmath2 since the number of plaquettes is the same as the number of local minima . to simplify , it is convenient to rotate the lattice by @xmath262 ( as shown in fig .
[ fig : plaque ] ) .
.,width=377 ] we then have a plaquette model where at each corner of a plaquette lives a random variable @xmath24 drawn from a uniform distribution over @xmath32 $ ] and we are interested in calculating the product in eq .
( [ plaq2 ] ) .
finally , the probability of the maximally packed configuration on the original lattice is obtained by averaging over the @xmath24 s with uniform distribution in @xmath32 $ ] @xmath263 where the angled brackets simply indicates integration over all the @xmath20 variables from @xmath21 to @xmath22 .
note that this plaquette model is very general and can be extended to any bipartite lattice . also , the right hand side of eq .
( [ plaq3 ] ) is actually the probability of just one of the two checkerboard configurations , so the true @xmath148 is actually twice this amount , but to avoid unnecessary complications we shall keep to this notation . as a simple example , for a @xmath22-d chain
, one can reproduce the result @xmath149 quite easily using the plaquette representation .
the number of plaquettes is clearly @xmath264 . in this case , eq .
( [ plaq3 ] ) gives @xmath265 where @xmath4 runs over every alternate sites of the original @xmath22-d chain .
the quantity on the r.h.s . of ( [ cd1 ] ) can be evaluated using a simple transfer matrix approach . defining a transfer matrix via , @xmath266 , we have from eq .
( [ cd1 ] ) , assuming a closed periodic chain @xmath267 \label{cd2}\ ] ] where @xmath268 is the trace .
the eigenvalue equation of the transfer matrix is @xmath269 where @xmath270 is an eigenfunction with eigenvalue @xmath219 . dividing the range of integration into @xmath271 $ ] and @xmath272 $ ] and differentiating twice
, one gets @xmath273 with the boundary conditions @xmath274 and @xmath275 where @xmath276 .
the solution is simply , @xmath277 where @xmath278 the largest eigenvalue is @xmath279 corresponding to @xmath280 .
thus for large @xmath264 , one gets from eq .
( [ cd2 ] ) @xmath281^{n/2}\sim [ \pi/2]^{-n } \label{cd6}\ ] ] reproducing the results obtained in the previous section .
another nontrivial solvable case of the plaquette model is on a two - leg ladder shown in fig .
[ fig : ladder ] .
it can be thought of as the first layer of the full @xmath150-d model shown on the right in fig . [
fig : plaque ] . at the center of each plaquette
lives a local minimum .
let @xmath2 be the number of plaquettes .
then the total number of sites in the lattice ( counting the minima at the centers ) is @xmath282 . the probability of the maximally packed configuration in eq .
( [ plaq2 ] ) can again be computed by defining the transfer matrix , @xmath283 where @xmath284 refers to the two corners on the left of a plaquette and @xmath285 refers to the two corners on the right of a plaquette .
then , @xmath267 \label{l1}\ ] ] the corresponding eigenvalue equation is now two dimensional @xmath286 which is considerably harder to solve compared to the @xmath22-d case . while one can solve the @xmath150-d eigenvalue equation directly , it is somewhat easier to first reduce it to an equivalent @xmath22-d eigenvalue problem by using the following trick .
let us first define the random variable @xmath287 at the center of each plaquette @xmath4 .
clearly @xmath288 where @xmath289 is the number of plaquettes . the random variables @xmath290 s are obviously correlated as two adjacent plaquettes will share two common random variables as shown in fig .
[ fig : ladder1 ] . at the center .
on the right we have two adjacent plaquettes with minima @xmath291 and @xmath292 at their respective centers .
the variables @xmath293 and @xmath294 are corrrelated as they share the common bond with the two elements @xmath253 and @xmath254.,width=529 ] to evaluate the average on the r.h.s of eq .
( [ l3 ] ) we need to know the joint distribution of the @xmath290 s which are correlated ( see fig . [
fig : ladder1 ] ) .
this joint distribution can be explicitly computed . to see this ,
let us first compute the cumulative distribution of one single @xmath91 variable , say @xmath295 . for simplicity ,
we denote @xmath296 , @xmath297 etc .
( see fig .
[ fig : ladder1 ] ) .
evidently @xmath298 where we have used the fact that @xmath251 , @xmath252 , @xmath253 and @xmath254 are all independent random variables drawn from the uniform distribution over @xmath32 $ ] .
next , let us consider the joint distribution of two consecutive @xmath91 s , say @xmath293 and @xmath294 .
let us denote the random variables at the corners by @xmath251 , @xmath252 , @xmath253 , @xmath254 , @xmath299 and @xmath300 as shown in fig .
[ fig : ladder1 ] .
then it is easy to see that @xmath301 ^ 2\,(1-y_2)^2 .
\label{l5}\end{aligned}\ ] ] in a similar way , one can construct the joint distribution of three adjacent plaquette minima @xmath302 ^ 2\ , \left[1-{\rm max}(y_2,y_3)\right]^2\ , ( 1-y_3)^2 . \label{l6}\ ] ] one can repeat the process above for higher number of adjacent plaquettes and one immediately sees the pattern for the full ladder .
let us denote @xmath303 then , for a ladder with open ends , we get @xmath304 ^ 2\,\right ] ( 1-y_m)^2 . \label{l7}\ ] ] for a ladder with closed ( periodic ) ends , this is even simpler @xmath305 ^ 2 .
\label{l8}\ ] ] where one identifies @xmath306 .
once we have the joint distribution , it is easy to rewrite the average on the r.h.s . of eq .
( [ l3 ] ) in terms of the joint distribution @xmath307 the latter identity can be easily derived using integration by parts .
let us , for simplicity , consider a plaquette with closed ends . then , substituting the joint distribution from eq .
( [ l8 ] ) into eq .
( [ l9 ] ) , we now have a one - dimensional multiple integral to peform @xmath308 ^ 2 .
\label{l10}\ ] ] using the identity @xmath309 and making a change of variable @xmath310 , the integral in eq .
( [ l10 ] ) simplifies further @xmath311 . \label{l11}\ ] ] this one dimensional integral can now be performed by a transfer matrix technique which operates in one dimension . defining @xmath312 we get , @xmath313 $ ] and the eigenvalue equation is given by @xmath314
thus we have managed to reduce a two - dimensional eigenvalue problem in eq .
( [ l2 ] ) into an equivalent but much simpler @xmath22-d eigenvalue problem in eq .
( [ l12 ] ) which can then be solved exactly .
to proceed , we first divide the range of integration in eq .
( [ l12 ] ) into @xmath271 $ ] and @xmath272 $ ] .
next we differentiate once to get @xmath315 note the boundary conditions that emerge from eqs.([l12 ] ) and ( [ l13 ] ) : @xmath274 and @xmath316 .
but in addition one also has to satisfy @xmath275 . dividing eq .
( [ l13 ] ) by @xmath20 and differentiating once more we get an ordinary second order differential equation @xmath317 with the boundary conditions : @xmath274 , @xmath316 and @xmath275 . the general solution to this equation , after a few changes of variables , can fortunately be obtained explicitly as a linear combination of two independent bessel functions . the boundary condition @xmath318 rules one of them out .
then , the most general solution satisfying @xmath318 can be written as @xmath319 where @xmath320 is the ordinary bessel function with index @xmath321 @xcite and @xmath322 is an arbitrary amplitude .
note that the other boundary condition at @xmath323 , namely @xmath324 is automatically satisfied by the solution in eq .
( [ l15 ] ) . this can be seen by using the small @xmath20 expansion of @xmath325 which indicates @xmath326 as @xmath327 .
hence @xmath324 . to determine the eigenvalue @xmath219
, we have to use the other nontrivial boundary condition at @xmath328 , namely @xmath275 .
this condition , substituted in eq .
( [ l15 ] ) , gives us an implicit equation for @xmath219 that looks a bit complicated @xmath329 however , a nice simplication occurs when one uses the identity @xcite , @xmath330 .
( [ l16 ] ) simply gives @xmath331 the bessel function oscillates on the positive axis of its argument , so each of its zeroes ( roots ) would give an eigenvalue @xmath219 .
however , for large @xmath2 , we are interested only in the largest eigenvalue which is then given by @xmath332 where @xmath333 is smallest positive root of @xmath334 .
the root @xmath333 is known @xcite , @xmath335 .
using @xmath336\sim \lambda^m$ ] for large @xmath2 and using @xmath289 , we finally get the exact result for the two - leg ladder @xmath337^{1/3}=1.57657\ldots \label{l19}\ ] ] when compared with the @xmath22-d result , @xmath338 , we see that @xmath75 changes by a very small amount as one goes from a chain to a ladder
. it would be interesting to see if one can extend these calculations to ladders with more than two legs @xcite and eventually to the full @xmath150-dimensional lattice .
now we consider the dependence of @xmath75 on dimension . for simplicity , we have focused on @xmath0-dimensional hypercubic lattices as these lattices are simple to parametrize and are bipartite , allowing us to use the @xmath0-dimensional analog of the framework described in sect .
[ subsect : plaquette ] .
our approach will be computational : we numerically estimate @xmath176 from an integral representation for lattices of increasing size and then try to extract the limit of large sizes .
to avoid boundary effects which slow down the convergence to this limit , we have used @xmath339 lattices having periodic boundary conditions in all directions .
the integral in eq .
( [ plaq3 ] ) is based on decomposing the lattice into even and odd sites and imposing the local minima to be say on the even sites .
its generalization to @xmath0 dimensions involves all the neighbors of a given even site @xmath9 , the corresponding `` star '' set of @xmath340 odd sites playing the role the plaquette had for the square lattice : @xmath341 here the average is over all values of the random variables belonging to stars ; these are uniform i.i.d . in @xmath32
$ ] . furthermore , as before
, we neglect the factor 2 in @xmath148 coming from the fact that the minima could have been taken to be on the odd sites .
the difficulty in computing these integrals is their high dimensionality . in monte carlo as used in most statistical physics applications , it is straightforward to use importance sampling methods ( e.g. , the metropolis algorithm ) to get expectation values of observables ; unfortunately here , the quantity to compute is the analog of the free energy and it is not directly accessible via such methods .
we have thus used a different approach that is of the `` variance reduction '' type .
it can be motivated as follows . in eq .
( [ eq : star ] ) we are to get the mean value of the integrand , sampling the random variables @xmath24 uniformly .
it is quite easy to see that the signal to noise ratio goes to zero exponentially with the number of lattice sites ; to counteract this , we sample the @xmath24 with a different density and then correct for this biased sampling . for this to be practical ,
we keep the @xmath24 as i.i.d .
variables , but optimize their individual distribution so as to maximize the signal to noise ratio .
let @xmath342 be the probability density used for sampling the @xmath24 .
for any such distribution , @xmath343^{1/2d } } \big \rangle_{\rho } .
\label{eq : starvr}\ ] ] the denominator corrects for the modified measure of the random variables , taking into account the fact that each odd site appears in @xmath340 stars . if @xmath344 is well chosen , the numerator and denominator of the integrand will fluctuate together so that their ratio has a reduced variance . for our purposes , we parametrized @xmath344 as follows : @xmath345 where @xmath346 and @xmath347 are arbitrary parameters while @xmath322 and @xmath348 are set so that @xmath344 is continuous and is a normalized probability density . for each dimension
, we adjust @xmath346 and @xmath347 to minimize the variance the integrand ; the signal to noise ratio still decreases exponentially with the number of lattice sites , but with a smaller exponent .
the motivation for this functional form is simple : the odd sites , all of which are maxima , have an a posteriori distribution that is strongly suppressed at low values of the random variable .
one last obstacle comes from the fact that the integrand typically takes on small values when one takes large lattices ; this is expected of course since the integral itself is becoming exponentially small . to keep track of values that are smaller than what can be represented by the machine coding , ( we used 96 bit representations of real numbers ) , we shifted multiplicatively each star term in the integrand and corrected for this shift when computing @xmath349 .
the set of these procedures then gave us values of @xmath350 with measurable statistical errors for a range of @xmath351 from which we extrapolated to the large @xmath3 limit .
in practice , we find that this strategy works very well in low dimensions .
for instance in dimension @xmath133 , we still have a very good precision for @xmath352 sites , and the large @xmath353 limit can be very reliably extracted , giving @xmath75 to better than 5 significant figures .
( of course , since the exact value is known at @xmath133 , this really only serves as a check of our procedures . ) in dimension @xmath354 , the method gives better than 4 significant figures for @xmath355 .
our estimates of @xmath356 as a function of @xmath357 are shown in fig .
( [ fig : gammaconvergence ] ) .
the convergence to the @xmath358 limit seems to follow a @xmath359 law , a property we also find for the higher dimensions investigated .
limit of the estimates of @xmath360 for the square lattice , @xmath361.,width=377 ] here we see that @xmath360 converges rapidly to a value close to 1.658 .
unfortunately the signal to noise ratio decreases with @xmath353 and with dimension ; furthermore the accessible range of @xmath353 decreases fast as @xmath0 increases ; because of this , we are able to extract @xmath75 reliably only for dimensions up to @xmath362 .
our results are summarized in table [ tab : numerics ] where the error estimates come from both statistical noise and uncertainties in the large @xmath353 extrapolation .
ccccccc @xmath363 & @xmath133 & @xmath354 & @xmath364 & @xmath365 & @xmath362 + + @xmath366 & 1.57082(6 ) & 1.6577(6 ) & 1.7152(10 ) & 1.761(5 ) & 1.806(10 ) + @xmath367 & 1.570796 & 1.854075 & 1.927621 & 1.956922 & 1.971464
we have been concerned with the statistical properties of @xmath2 , the number of local minima in a random energy landscape .
the low order moments of @xmath2 can easily be obtained ; higher order moments could be computed by automated counting of graphs .
atypical _ values of @xmath2 follow a large deviation principle as given in eq .
( [ ld1 ] ) ; we have been able to compute this function @xmath57 in one dimension analytically ; it diverges as @xmath126 and goes to a finite limit when @xmath2 reaches its maximum value where one half of the lattice sites are local minima .
this then led us to consider the limit of maximum packing ( @xmath2 takes on its largest possible value ) for more general lattices .
we focused on bipartite lattices where up to half of the sites can be local minima .
we derived analytically the probability of this maximum packing @xmath368 for the cayley tree and for a two - leg ladder .
we then tackled @xmath0-dimensional hypercubic lattices by computational techniques .
for all these lattices , it is easy to see that @xmath369 simply by forcing the @xmath24 on the even ( odd ) sites to be less ( greater ) than @xmath370 ; this immediately leads to @xmath371 for all dimension .
furthermore when @xmath0 becomes large , it will be very rare to have maximum packing if the @xmath24 do not very nearly satisfy this even - odd pattern so one expects @xmath372 to tend towards 2 in the large @xmath0 limit . given this large @xmath0 limit , it is natural to ask how @xmath246 is approached . in fig .
( [ fig : gamma ] ) we show that the cayley tree case follows a very clear power correction law which can be derived analytically as being @xmath373 .
the case of the @xmath0-dimensional lattice is less clear but is compatible with a @xmath374 law .
the cayley tree thus provides an approximation to the hypercubic case but not a very accurate one ; this is probably because the nature of the correlations from site to site of the @xmath24 ( given that one has a maximally packed configuration ) are quite different when considering the tree rather than the hypercube .
our work can be extended in several ways .
( i ) we computed the exact large deviation function in one dimension , but this function can also be determined for the cayley tree @xcite .
( ii ) one can also introduce a chemical potential @xmath91 for each minimum and consider the thermodynamics of minima as a function of @xmath91 ; recent numerical results by derrida in @xmath150-d @xcite indicate the presence of a phase transition with @xmath91 .
( iii ) our variables @xmath24 were i.i.d .
random variables ; does a large deviation principle still hold if these variables have short range correlations ?
one expects so .
( iv ) how do all these statistical properties generalize if one asks for minima within a given energy range ? finally , it would also be of interest to understand the similarities and differences between the statistics of local minima in continuous random energy landscapes @xcite and in lattice models as presented here ; the energies could be random or correlated as in the sinai problem .
this work was supported by the eec s fp6 information society technologies programme under contract ist-001935 , evergrow ( www.evergrow.org ) , and by the eec s hpp under contract hprn - ct-2002 - 00307 ( dyglagemem ) .
s.m . wants to thank the hospitality of the isaac newton institute , cambridge ( uk ) where this work was completed .
we thank our colleagues a.j .
bray , b. derrida , d. dhar , m.r .
evans , k. mallick , s. nechaev , g. oshanin and p. sollich for stimulating discussions . | we consider random energy landscapes constructed from @xmath0-dimensional lattices or trees .
the distribution of the number of local minima in such landscapes follows a large deviation principle and we derive the associated law exactly for dimension 1 . also of interest
is the probability of the maximum possible number of minima ; this probability scales exponentially with the number of sites .
we calculate analytically the corresponding exponent for the cayley tree and the two - leg ladder ; for 2 to 5 dimensional hypercubic lattices , we compute the exponent numerically and compare to the cayley tree case . for many materials that are glassy @xcite local minima of the energy ( or of the free energy ) trap the system for long times , leading to subtle equilibrium and out of equilibrium properties .
energy landscapes provide a simple conceptual framework for modeling these systems but in fact their use goes much beyond that .
for instance the vacua of string theories are expected to proliferate enormously , and the problem of estimating the number of local minima @xcite of string energy landscapes is still an open problem .
other examples include rugged ( random ) energy landscapes in evolutionary biology @xcite , quantum cosmology @xcite , manifolds in random media @xcite , glassy systems @xcite random potentials @xcite and the associated problems in random matrix theory @xcite .
typically these systems consider a particle , a configuration of particles , or even a manifold , subject to a random potential .
one extreme case for its visual simplicity is that of a point particle in a random gaussian potential ; on the opposite extreme are energy landscapes associated with the configuration of a many body system .
examples in this last category include : ( 1 ) the @xmath1-spin glass model @xcite , which in the limit of large @xmath1 reduces to the far simpler random energy model of derrida @xcite ; ( 2 ) atomic clusters @xcite and other glassy systems with no quenched disorder @xcite ; ( 3 ) random manifolds in random media @xcite .
for any landscape , it is desirable to know the statistical properties of the minima ( or more generally of the saddle points ) .
a quantity frequently considered is the expected number of minima as a function of their energy @xcite .
it may also be of interest to consider how the set of minima are organized topologically , e.g. , whether the barrier tree @xcite is ultrametric .
our focus here is to better understand the statistics of the total number of local minima in random energy landscapes ; our underlying space is a regular ( euclidean ) lattice on each site of which resides a random energy .
let @xmath2 denote the total number of local minima for given values of the energies on each site .
evidently @xmath2 is a random variable as it varies from one realization of the landscape to another .
we are interested in the statistics of @xmath2 as a function of the number @xmath3 of lattice sites .
similar questions were studied recently in the context of random permutations @xcite , ballistic deposition @xcite , and in simple models of glasses @xcite . in this paper
, we provide a number of analytical results on the distribution of the total number of local minima for random energy landscapes on several lattices ; the moments of @xmath2 are easily derived , so our focus concerns mainly the probability of _ large deviations _ of @xmath2 from its mean value , i.e. , the probabilities of atypical configurations .
furthermore we ask what is the probability of @xmath2 being at its maximum , corresponding to the limit where the minima are maximally packed on the lattice .
the paper is organized as follows . in sect .
[ sec : model ] we specify the model and show that the statistical properties of the minima are independent of the individual distribution of energies as long as these energies are independent from site to site and are drawn from a continuous distribution . in sect .
[ sec : generalproperties ] we cover some of the simplest properties of the statistics of @xmath2 and formulate the large deviation principle .
then we derive the closed form expression for the large deviation function in the case of the one dimensional lattice in sect .
[ sec : onedimension ] .
the focus of the rest of the paper is the maximum packing problem . in sect .
[ sec : maxpack ] we determine the probability of maximum packings for several solvable cases , namely cayley trees and a two - leg ladder .
the case of @xmath0-dimensional hypercubic lattices is then treated by numerical computation in sect .
[ sec : numerics ] .
finally , these different results are discussed in sect .
[ sec : conclusions ] and some closing remarks are given . |
low energy effective theories derived from the ns - ns sector of string theories contain the gravity , @xmath0 , the dilation , @xmath1 , and the antisymmetric tensor field , @xmath2 .
the existence of the last degree of freedom leads to intriguing implications in string cosmology @xcite . in four spacetime dimensions ,
the massless antisymmetric tensor field is dual to the a pseudo - scalar ( axion ) field @xcite , and the axion
dilaton system is known to develop an unobserved anisotropy in our universe , which can be diluted away at late times only in a contracting universe @xcite .
such a disastrous cosmological situation can be resolved also in string theory which contains another dynamical object called d - brane @xcite .
the gauge invariance on the d - brane is maintained through the coupling of the gauge field strength to the antisymmetric tensor field @xcite .
then , the effective action derived on the d - brane describes the antisymmetric tensor field as a massive and self - interacting 2-form field .
cosmological evolution of such a tensor field has been investigated in ref .
@xcite assuming that the dilaton is fixed to a reasonable value .
although the time - dependent magnetic @xmath3 field existing in the early universe develops an anisotropy in the universe , it was realized that the matter - like behavior of the @xmath3 field ( b - matter ) ensures a dilution of the anisotropy at late times and thus the isotropy is recovered in reasonable cosmological scenarios @xcite . in such sense
the effect of antisymmetric tensor field on the d - brane is distinguished from that of field strength of the u(1 ) gauge field @xcite . in this paper
, we investigate the cosmological evolution of the b - matter dilaton system in our universe which is assumed to be imbedded in the d - brane .
the usual string cosmology with the dilaton suffers from the notorious runaway problem , which is also troublesome in the d - brane universe . in our study ,
the dilaton obtains two exponential potential terms due to the curvature of extra dimensions @xmath4 , and the d - brane tension ( the mass term of the b - matter ) @xmath5 .
it is interesting to observe that the dilaton can be stabilized for negative @xmath4 @xcite which , however , leads to a contracting universe due to the effective negative cosmological constant in our universe .
when @xmath4 is positive , the b - matter dominance will be overturned by the @xmath4 dominance as the dilaton runs away to the negative infinity . as a consequence of this , the initial anisotropy driven by the b - matter
can also be diluted away at late times . for a realistic low energy
effective theory , string theory must be endowed with a certain mechanism generating an appropriate vacuum expectation value for the dilaton . in such a situation , the dilaton is expected to be stabilized at some stage of the cosmological evolution affecting the dynamics of the antisymmetric tensor field .
taking an example of the dilaton stabilization , we will also examine the cosmological evolution of the b - matter and the dilaton in which the essential features of ref .
@xcite are reproduced .
this paper is organized as follows . in section [ sec:2 ]
, we describe the low energy effective action of the d - brane universe and the corresponding field equations . before our main discussion ,
section [ sec:3 ] is devoted to presenting homogeneous solutions in flat spacetime illustrating some interesting features of the b - matter dilaton system in a simple way . in section [ sec:4 ] , we find semi - analytic and numerical cosmological solutions to observe an intriguing interplay of the curvature @xmath4 and the d - brane tension @xmath5 . in section [ sec:5 ]
, we consider the evolution of the anisotropic universe with the dilaton stabilization which leads to a satisfactory cosmology of the d - brane universe .
we conclude in section [ sec : conc ] .
the main idea of the d - brane world is that we reside on a d@xmath6-brane imbedded in 10 ( or 11 ) dimensional spacetime with extra - dimensions compactified .
the bosonic ns - ns sector of the d - brane world consists of the u(1 ) gauge field @xmath7 living on the d@xmath6-brane and the bulk degrees including the graviton @xmath0 , the dilaton @xmath1 , and the antisymmetric tensor field of rank - two @xmath2 . in the presence of the brane
, the gauge invariance of @xmath2 is restored through its coupling to a u(1 ) gauge field @xmath7 and the gauge invariant field strength is @xcite @xmath8 even though we are assumed to live on the d - brane , we adopt the conventional compactification in the sense that the extra dimensions are compact and stabilized , and thus static . in the presence of the d - brane ,
this would in general require an additional setup like additional branes and fluxes along the extra dimensions @xcite , and the working of non - perturbative effects @xcite .
the extra dimensions are warped due to branes and fluxes .
the warping of extra dimensions gives a warp factor in the definition of the four - dimensional planck scale @xcite , and also induces the potential for the dilaton @xcite as described below .
if the fluxes in the extra dimensions significantly affect the compactification , the existence of the fluxes on the d - brane seems also natural and it is intriguing to tackle their effect in the early universe
. however , simultaneous consideration of the fluxes along both the extra dimensions and the d - brane lets the computation almost intractable .
therefore , the conventional compactification is an appropriate setup at the present stage , and then we take the following four - dimensional effective action of the bosonic sector in the string frame @xcite @xmath9,\end{aligned}\ ] ] where the tilde denotes the string frame quantity , @xmath10}$ ] and @xmath11 with @xmath12 . the parameter @xmath5 is defined by @xmath13 where @xmath14 is the effective brane tension .
note that we omitted the chern - simons like term assuming the trivial r - r background .
if we assume that the six extra - dimensions are compactified with a common radius @xmath15 , one finds @xmath16 where @xmath17 and @xmath18 is the four - dimensional planck mass .
the qualitative features of our results do not depend on specific values of @xmath6 .
thus , we take @xmath19 for simplicity .
the @xmath4 term comes from the scalar curvature or the condensate @xmath20 of extra dimensions integrated over the whole extra dimensions .
the action , and thus also field equations , can be written in a more familiar form in the einstein metric , which is defined by @xmath21 we will work in this metric from now on . in terms of einstein metric , the action becomes @xmath22.\end{aligned}\ ] ] the field equations derived from the action ( [ action - e ] ) are @xmath23 @xmath24 @xmath25 where the dilaton potential is @xmath26 , \label{dilaton - potential}\end{aligned}\ ] ] and the energy - momentum tensor is given by @xmath27 in the subsequent sections , we examine the equations of motion ( [ b - eq1])([e - eq1 ] ) and find homogeneous solutions .
cosmological implication of the d - brane universe is of our main interest , including stabilization of the dilaton and condensation of the antisymmetric tensor field .
for the sake of the simplicity , let us first look into the dynamics of our system in flat spacetime . with @xmath28 ,
the einstein equations ( [ e - eq1 ] ) become simple and gauge invariance of the field variables @xmath29 and @xmath30 which directly appear in the expression of energy - momentum ( [ tb ] ) allows a homogeneous configuration @xmath31 for homogeneous configuration ( [ hom ] ) , @xmath32-components ( _ electric _ components ) of the equation of the antisymmetric tensor field ( [ b - eq1 ] ) reduce to an algebraic equation @xmath33 @xmath34-components ( _ magnetic _ components ) give @xmath35 where the overdot denotes differentiation with respect to the rescaled dimensionless time @xmath36 these equations ( [ 0i])([ij ] ) tell us that the _ magnetic _ components @xmath37 are _ dynamical _ but the _ electric _ components are determined by the _ constraint _ equation ( [ 0i ] ) .
dynamics of the dilaton is governed by ( [ p - eq1 ] ) which becomes @xmath38 where the dimensionless parameter @xmath39 is defined by @xmath40 the constraint equation ( [ 0i ] ) forces its numerator to vanish except a trivial solution @xmath41 , however vanishing numerator allows only a _ vanishing
_ electric field solution @xmath42 .
then the equations for the magnetic components ( [ ij ] ) and the dilaton ( [ de ] ) reduce to @xmath43 in the following , we will find solutions of ( [ m1 ] ) and ( [ m2 ] ) to see how the dilaton and the magnetic component behave .
when the magnetic components of the antisymmetric tensor field @xmath45 vanish , we have only the dilaton equation from ( [ m2 ] ) @xmath46 where the dilaton potential @xmath47 is given by the sum of the curvature term and the brane tension term @xmath48 as shown in figure [ fig1 ] . .
for @xmath49 , @xmath47 monotonically increases ( dashed line ) . for @xmath50 , @xmath47 has a minimum at @xmath51 ( solid line ) . ] for @xmath49 , the dilaton potential @xmath47 monotonically increases , so the dilaton finally rolls to the minimum @xmath52 as @xmath53 . for @xmath50 ,
the dilaton potential @xmath47 has the minimum value @xmath54 at @xmath55 . if @xmath56 , the dilaton is stuck at @xmath57 .
when @xmath58 , the dilaton oscillates around the minimum .
therefore , the dilaton is stabilized with mass @xmath59 , and due to exponential potential terms , its mass can be much different from the vacuum expectation value of the dilaton @xmath60 .
as @xmath39 increases , the dilaton mass rapidly grows .
when @xmath61 , the dilaton again rolls to negative infinity .
it means that the dilaton is easily destabilized by its fluctuations , if the dilaton mass becomes too small .
though the negative vacuum energy @xmath62 does not affect the dynamics of the dilaton and the antisymmetric tensor field in flat spacetime , its coupling to gravity leads to singular cosmological evolution as we shall see in section [ sec:4 ] .
therefore , for the proper description of cosmology , some additional term is needed to adjust the value of @xmath63 to vanish as we shall discuss in section [ sec:5 ] .
the runaway or stabilized behavior of the dilaton remains essentially unchanged even with @xmath64 although the dependence on the initial conditions alters as we will see in the next subsection . without loss of generality ,
let us assume that the direction of the magnetic component of antisymmetric tensor field is fixed as @xmath65 .
although the equations of motion ( [ m1])([m2 ] ) are still complicated even under this assumption , the presence of the magnetic components do not alter the story of the dilaton much .
first , let us consider a simple case of zero brane tension ( @xmath66 ) , which is analytically tractable .
then in ( [ m1])([m2 ] ) , only the @xmath39-dependent term survives on the right - hand side .
( for this case , we will use the original @xmath4 and time @xmath67 since @xmath66 . ) the solutions to these equations are @xmath68 + \ln c_1 \,,\ ] ] and @xmath69 where @xmath70 , @xmath71 , and @xmath72 are integration constants , and the time has been rescaled to @xmath73 by absorbing another constant @xmath74 .
the solutions are still remaining valid under @xmath75 , i.e. , under the change of the sign of @xmath72 .
therefore , we can fix the time @xmath76 to flow to the positive direction and @xmath77 .
the solutions approach their asymptotic configurations at large @xmath76 , @xmath78 we observe that the dilaton rolls linearly in time to negative infinity , and the antisymmetric tensor field condensates to a constant @xmath79 which depends on @xmath4 .
there are two topologically distinct condensation processes depending on the signature of @xmath70 which is to be set by the initial conditions ; the change in @xmath4 contributes to @xmath80 oppositely .
with the brane tension term turned on , we numerically study due to the complexity of field equations .
the solutions are classified into two classes as in the pure dilaton case in section [ subsec:3.a ] , depending on the signature of @xmath39 . for @xmath49 ,
the dilaton approaches negative infinity irrespective of initial conditions , @xmath81 and @xmath82 .
the condensed value of the magnetic component deviates from the initial value in the beginning , and then stabilizes to a constant as shown in figure [ fig2 ] .
we observed from the numerical results that there exist two branches of the magnetic condensation depending on the initial conditions as discussed above , but we show only one branch in the figure . like in the @xmath66 case , once the magnetic component is condensed , the condensation survives as a constant value .
it is different from the case without the dilaton , in which the magnetic component permanently oscillates in flat spacetime @xcite . and @xmath3 in the einstein frame for @xmath83 ( solid lines ) , @xmath84 ( dashed lines ) , and @xmath85 ( dotted lines )
. the initial conditions are @xmath86 and @xmath87 .
the configurations are very similar regardless of @xmath88.,height=226 ] and @xmath3 in the einstein frame for @xmath89 , @xmath87 , and @xmath90 the dilaton stabilization is observed as well as the periodically waving oscillation of @xmath3 . ] when @xmath39 is negative , there are two classes of solutions , depending on the initial value of the dilaton @xmath81 ( or equivalently the energy density ) for a fixed initial magnetic condensation @xmath82 . for large @xmath81 or @xmath82 ,
the configurations are almost the same as those for nonnegative @xmath39 ( see figure [ fig2 ] ) .
for small enough @xmath81 and @xmath82 , the dilaton shows the oscillating behavior as expected from the dilaton potential in figure [ fig1 ] .
according to the dilaton oscillation , the magnetic component is also oscillating around a condensed value ( see figure [ fig3 ] ) .
if we involve the radiation of perturbative modes of the dilaton and the antisymmetric tensor field , which is missing in this classical configuration , the classical solution will damp to a stabilized value of the dilaton and the final condensed value of the magnetic component .
in this section we study cosmological solutions of the d - brane universe in the presence of both the dilaton and the antisymmetric tensor field . as we have done in the previous paper @xcite
, we assume spatially homogeneous configurations for the antisymmetric tensor field and the dilaton , and look for the time evolution of these fields and the expansion of the universe . the non - vanishing homogeneous antisymmetric tensor field , in general , implies the anisotropic universe . to consider the simplest form of anisotropic cosmology ,
we take only a single magnetic component of @xmath91 to be nonzero , namely @xmath92 and @xmath93 .
then the metric consistent with this choice of field configuration is of bianchi type i @xmath94 then , in the einstein frame , the field equation for @xmath3 ( [ b - eq1 ] ) reads @xmath95 the dilaton - field equation ( [ p - eq1 ] ) is @xmath96 and einstein equations ( [ e - eq1 ] ) are @xmath97 where @xmath98 it is more convenient to employ the dimensionless time variable @xmath99 of ( [ ti ] ) and to introduce the variables @xmath100 and @xmath101 defined by @xmath102 then the above equations ( [ eq = beq])([eq = g33 ] ) are rewritten as @xmath103b & = & 0,\end{aligned}\ ] ] @xmath104 @xmath105 where , by dividing @xmath106 from each @xmath107 in eqs .
( [ br1])([br2 ] ) , we have @xmath108 ^ 2,\quad \rho_\lambda=\lambda e^{2\phi } , & \nonumber\\&\displaystyle \rho_b=\frac12e^{3\phi}(1+b^2)^{1/2},\quad \tilde\rho_b=\frac12e^{3\phi}(1+b^2)^{-1/2 } , & \end{aligned}\ ] ] where @xmath39 is defined in ( [ co ] ) . in the subsequent subsections ,
we look for the solutions to the above field equations for various cases beginning with some simple solutions .
the dilaton in the system of equations ( [ eq - b])([eq - a3 ] ) has the exponential potential up to the correction due to the antisymmetric tensor field .
it is well - known that the scalar field with the exponential potential possesses the scaling solution in which the energy density of the scalar field mimics the background fluid energy density @xcite .
this scaling solution is also an attractor , so that the late time behavior of the solutions are universal irrespective of initial conditions .
this is an attractive property of the exponential potential . for the potential of the form @xmath109
, there is an attractor solution @xmath110,\quad \alpha=\left(\frac{2}{\beta}\right)^2\ln t,\ ] ] for @xmath111 .
the scale factor obeys the power - law time - dependence , implying that the rolling of @xmath1 constitutes the matter having an equation of state @xmath112 where @xmath113 varies from @xmath114 to @xmath115 for the aforementioned range of @xmath116 .
we found this type of particular solutions of the eqs .
( [ eq - b])([eq - a3 ] ) , which can be found when we have a single exponential term in the potential , that is , for the case of @xmath117 and for the case of @xmath66 . for both cases
we start from an ansatz of the form @xmath118 where we suppressed constant terms in @xmath119 and @xmath120 which correspond to simple rescaling of coordinates .
for the case of @xmath117 , we obtain two distinguished solutions @xmath121 and @xmath122 the first solution is nothing but the solution ( [ sol - a ] ) with @xmath123 .
the second solution has the non - vanishing antisymmetric tensor field . for the case of @xmath66
, we introduce a new rescaled dimensionless time variable @xmath124 instead of @xmath125 of ( [ ti ] ) and then get the continuous set of solutions from eqs .
( [ eq = beq])([eq = g33 ] ) @xmath126 @xmath127 and @xmath128 are same as those in eq .
( [ sol - a ] ) with @xmath129 , while we have the non - vanishing @xmath3 field condensate .
these solutions are the solutions to the eqs .
( [ eq - b])([eq - a3 ] ) for the specific initial conditions .
however , the importance of these solutions , as noted in the first paragraph , arises from the fact that they are attractors , which means that after enough time the solutions with different initial conditions approach these solutions .
we will confirm this through numerical analysis in the next subsection .
the solution ( [ sol-1 ] ) applies for the brane tension dominated case where the dilaton potential is approximated by @xmath130 from eq .
( [ pu ] ) . the evolution of the dilaton under this potential produces matter with the equation of state @xmath131 .
once the antisymmetric tensor field is turned on , the anisotropy appears as in the solution ( [ sol-2 ] ) .
the measure of anisotropy is @xmath132 this result is contrasted with that in ref .
@xcite where the dilaton is assumed to be stabilized .
the rolling of the dilaton makes the difference .
it affects the dynamics of @xmath3 field in such a way that @xmath101 remains constant at @xmath133 instead of oscillating about the potential minimum @xmath134 and the anisotropy is maintained .
let us turn to the third solution ( [ sol-3 ] ) .
it is relevant when the dilaton potential arising from the curvature of extra dimensions , @xmath135 , dominates over other contributions . in our scheme
this happens as the dilaton rolls down the potential .
when @xmath101 vanishes , the transition point at which @xmath136 starts to dominate over @xmath137 is at @xmath138 .
thus , this solution describes the late time behavior of all the solutions with various initial conditions when @xmath4 is positive .
it is very intriguing since we achieve the isotropic universe in the end .
the rolling of dilaton makes the brane tension term which causes the anisotropy when we have non - vanishing @xmath3 field less important than the extra dimension curvature term which recovers the isotropy .
the rolling of the dilaton under the potential @xmath136 now forms the matter having the equation of state @xmath139 , thus giving marginal inflation .
we have the second order differential equations for four variables @xmath140 , @xmath141 , @xmath142 , @xmath143 .
thus we need eight initial values @xmath81 , @xmath144 , @xmath82 , @xmath145 , @xmath146 , @xmath147 , @xmath148 , @xmath149 to specify the solution . among these
, @xmath150 can always be set to zero by coordinate rescaling .
@xmath151 must obey the constraint equation ( [ eq - constraint ] ) , but this does not fix the ratio @xmath152 . we choose the isotropic universe as a natural initial condition which leads to @xmath153 .
since the dilaton potential is composed of exponential terms , the shift of the dilaton field by a constant can be traded for the redefinition of mass scale .
we use this property to take the initial value of the dilaton to be zero without loss of generality . in our numerical analysis , we take the dimensionless time variable as @xmath154 where @xmath155 and use the variable @xmath156 with its initial value @xmath157 .
this means that the proper time scale for the cosmological evolution is not @xmath158 , but @xmath159 . for convenience s sake , we take the initial time as @xmath160 .
the other mass scale @xmath4 is also affected by this shift and we can treat it by replacing the parameter @xmath39 with @xmath161 .
now we need three initial values @xmath144 , @xmath82 , and @xmath162 , to fix the functional form of the solution . the initial values @xmath163 and @xmath164 are related to @xmath82 and @xmath145 by @xmath165 and @xmath166 .
if we assume @xmath168 , we can study the role of the dilaton interacting with gravity more clearly as was also true in the subsection [ subsec:3.a ] .
we keep the brane term with the tension @xmath5 .
the spacetime is now isotropic .
the parameter @xmath4 in the string - frame action , which can be interpreted as the curvature of the compact internal manifold , plays a very important role , and the solutions are topologically different depending on the signature of @xmath4 . when @xmath117 , we have an attractor solution ( [ sol-1 ] ) .
since we consider the dilaton only , the only relevant initial condition is @xmath144 .
the initial condition for the attractor solution is @xmath169 in the setup described in the previous subsection .
for a few other values of @xmath144 , we plot the numerical solutions in figure [ fig4 ] . regardless of the initial value @xmath144 ,
all the solutions approach universally the attractor solution ( up to rescaling for the scale factor ) after some time . and @xmath117 with initial value @xmath169 ( solid curve , the attractor ) , @xmath170 ( dotted curve ) , @xmath171 ( dashed curve ) and @xmath172 ( dash - dotted curve ) , respectively . for the evolution of the scale factor , we plot @xmath173 for it becomes constant @xmath174 when the scale factor behaves as @xmath175.,title="fig:",height=245 ] and @xmath117 with initial value @xmath169 ( solid curve , the attractor ) , @xmath170 ( dotted curve ) , @xmath171 ( dashed curve ) and @xmath172 ( dash - dotted curve ) , respectively
for the evolution of the scale factor , we plot @xmath173 for it becomes constant @xmath174 when the scale factor behaves as @xmath175.,title="fig:",height=245 ] when @xmath176 , the evolution is divided into two stages .
when @xmath1 is large so that @xmath177 is much larger than @xmath178 , the solution approaches the attractor ( [ sol-1 ] ) . as @xmath1 rolls down to around @xmath138 , @xmath179 becomes larger than @xmath180 .
after that point the solution approaches the attractor ( [ sol-3 ] ) with @xmath134 . in figure
[ fig5 ] , numerical solutions show this two stage evolution explicitly . and @xmath176 .
when we set the parameter @xmath181 , the curves stand for the solutions with the initial value @xmath169 ( solid curve ) and @xmath170 ( dotted curve ) , @xmath171 ( dashed curve ) and @xmath172 ( dash - dotted curve ) , respectively.,title="fig:",height=245 ] and @xmath176 .
when we set the parameter @xmath181 , the curves stand for the solutions with the initial value @xmath169 ( solid curve ) and @xmath170 ( dotted curve ) , @xmath171 ( dashed curve ) and @xmath172 ( dash - dotted curve ) , respectively.,title="fig:",height=245 ] since @xmath4 measures the curvature of the extra dimensions , its value can be negative . when @xmath182 , the dilaton potential has the global minimum at @xmath183 with negative cosmological constant @xmath184 as discussed in the subsection [ subsec:3.a ] .
the disastrous consequence of this negative energy minimum is that the scale factor collapses in the end .
this can not be avoided because the dilaton rolls down toward the minimum and stays where the potential is negative while the kinetic energy is diluted by the expansion , thus the total energy becomes negative and it derives the universe to collapse .
it is different from the case of flat spacetime in the subsection [ subsec:3.a ] , where the dilaton experiences either the permanent oscillation around @xmath185 for negative initial energy density or the attractor solution for nonnegative initial energy density .
now , we turn on the antisymmetric tensor field along the @xmath187 direction , @xmath188 .
the spacetime becomes anisotropic , @xmath189 , in general .
the solutions are classified again by the signature of @xmath4 . and @xmath117 .
the initial conditions are as follow .
solid line : @xmath190 , @xmath191 , @xmath192 ; dotted line : @xmath193 , @xmath194 , @xmath192 ; dashed line : @xmath193 , @xmath195 , @xmath192 ; dash - dotted line : @xmath169 , @xmath194 , @xmath192 .
the lower - right panel shows the evolution of each component of energy density for initial conditions @xmath190 , @xmath191 , @xmath192.,title="fig:",height=245 ] and @xmath117 .
the initial conditions are as follow .
solid line : @xmath190 , @xmath191 , @xmath192 ; dotted line : @xmath193 , @xmath194 , @xmath192 ; dashed line : @xmath193 , @xmath195 , @xmath192 ; dash - dotted line : @xmath169 , @xmath194 , @xmath192 .
the lower - right panel shows the evolution of each component of energy density for initial conditions @xmath190 , @xmath191 , @xmath192.,title="fig:",height=245 ] + and @xmath117 .
the initial conditions are as follow .
solid line : @xmath190 , @xmath191 , @xmath192 ; dotted line : @xmath193 , @xmath194 , @xmath192 ; dashed line : @xmath193 , @xmath195 , @xmath192 ; dash - dotted line : @xmath169 , @xmath194 , @xmath192 . the lower - right panel shows the evolution of each component of energy density for initial conditions @xmath190 , @xmath191 , @xmath192.,title="fig:",height=245 ] and @xmath117 .
the initial conditions are as follow .
solid line : @xmath190 , @xmath191 , @xmath192 ; dotted line :
@xmath193 , @xmath194 , @xmath192 ; dashed line : @xmath193 , @xmath195 , @xmath192 ; dash - dotted line : @xmath169 , @xmath194 , @xmath192 .
the lower - right panel shows the evolution of each component of energy density for initial conditions @xmath190 ,
@xmath191 , @xmath192.,title="fig:",height=245 ] for @xmath117 , we have an attractor ( [ sol-2 ] ) under the initial conditions of @xmath190 , @xmath196 , and @xmath197 . in figure [ fig6 ] , we plotted numerical solutions for a few different initial conditions in addition to the attractor .
it is confirmed again that all the solutions approach the attractor as the time goes on .
the final value of @xmath101 is either @xmath198 or @xmath199 depending on the initial conditions . in the lower - right panel in figure [ fig6 ]
we provide the evolution of each component of energy density .
the kinetic energy of the dilaton @xmath200 catches up the potential energy @xmath180 and the ratio of them becomes constant .
this is a characteristic feature of the scaling solution @xcite .
the kinetic energy of @xmath3 field is kept much smaller than both of them , but the anisotropy is still maintained due to the difference between @xmath180 and @xmath201 . and @xmath176 . when @xmath181 , the solutions of four initial conditions are given : @xmath190 , @xmath191 , @xmath192 for the solid curves , @xmath193 , @xmath194 , @xmath192 for the dotted curves , @xmath193 , @xmath195 , @xmath192 for the dashed curves , and @xmath169 , @xmath194 , @xmath192 for the dash - dotted curves .
the lower - right panel shows the evolution of each component of energy density for initial conditions @xmath169 , @xmath194 , @xmath192.,title="fig:",height=245 ] and @xmath176 . when @xmath181 , the solutions of four initial conditions are given : @xmath190 , @xmath191 , @xmath192 for the solid curves , @xmath193 , @xmath194 , @xmath192 for the dotted curves , @xmath193 , @xmath195 , @xmath192 for the dashed curves , and @xmath169 , @xmath194 , @xmath192 for the dash - dotted curves .
the lower - right panel shows the evolution of each component of energy density for initial conditions @xmath169 , @xmath194 , @xmath192.,title="fig:",height=245 ] + and @xmath176 .
when @xmath181 , the solutions of four initial conditions are given : @xmath190 , @xmath191 , @xmath192 for the solid curves , @xmath193 , @xmath194 , @xmath192 for the dotted curves , @xmath193 , @xmath195 , @xmath192 for the dashed curves , and @xmath169 , @xmath194 , @xmath192 for the dash - dotted curves .
the lower - right panel shows the evolution of each component of energy density for initial conditions @xmath169 , @xmath194 , @xmath192.,title="fig:",height=245 ] and @xmath176 . when @xmath181 , the solutions of four initial conditions are given : @xmath190 , @xmath191 , @xmath192 for the solid curves , @xmath193 , @xmath194 , @xmath192 for the dotted curves , @xmath193 , @xmath195 , @xmath192 for the dashed curves , and @xmath169 , @xmath194 , @xmath192 for the dash - dotted curves .
the lower - right panel shows the evolution of each component of energy density for initial conditions @xmath169 , @xmath194 , @xmath192.,title="fig:",height=245 ] for @xmath176 , the evolution is again divided into two stages as in the @xmath167 case . in the first stage where @xmath180 is dominant
, the solution approaches an attractor ( [ sol-2 ] ) . in the second stage where @xmath179 is dominant
, it approaches another attractor ( [ sol-3 ] ) .
thus the universe recovers the isotropy .
the final value of @xmath101 is a certain constant which is determined by initial conditions and can differ from @xmath202 .
numerical solutions for a few initial conditions are shown in figure [ fig7 ] .
the kinetic energy of the dilaton @xmath200 now catches up the potential energy @xmath180 in the first stage and @xmath179 in the second stage .
the ratios , @xmath203 and @xmath204 , approach constants in each stage . the kinetic energy of @xmath3 field is kept much smaller as in @xmath117 case . for @xmath182 ,
the solution becomes singular as in the @xmath167 case .
here we skip the description of such singular solutions which are not suitable for the evolution of our universe .
the vacuum expectation value of the dilaton determines both the gauge and gravitational coupling constants of the low energy effective theory .
therefore , the dilaton must be stabilized at some stage of the evolution for the action ( [ action - e ] ) to have something to do with the reality . in this section ,
we study the cosmological evolution when the dilaton is stabilized . as for the correct mechanism of dilaton stabilization ,
the consensus has not been made yet .
our goal here is to illustrate an example of the dilaton stabilization and look into the effect of it on the dynamics of @xmath3 field and the cosmological evolution , since the overall features of which are insensitive to the detailed mechanism of stabilization .
our starting point is the dilaton potential ( [ dilaton - potential ] ) .
this potential possesses the minimum for @xmath182 , but the value of the potential at the minimum is negative and need to be set to zero by fine - tuning of the constant shift .
however , the constant shift of the potential has no motivation in the context of string theory .
instead , we introduce a term @xmath205 in the potential , which can arise from the effect of various form field fluxes in extra dimensions @xcite .
thus , our potential for the dilaton for @xmath167 looks like @xmath206 this potential has a global minimum for any value of @xmath4 and @xmath207 . to have sensible cosmology ,
the potential at the minimum must be zero . for @xmath182 , this can be done through a fine - tuning of the parameters in the potential @xmath208 where @xmath209
. then the minimum is located at @xmath210 the shape of this fine - tuned potential for @xmath211 is shown in figure [ fig8 ] .
now the equations ( [ eq - phi])([eq - a3 ] ) are modified accordingly @xmath212 where @xmath213 , while the equation ( [ eq - b ] ) for @xmath101 is not changed . with the stabilizing potential for the dilaton ,
the cosmological evolution is completely changed .
the dilaton and the antisymmetric tensor field rapidly come to the oscillation about the potential minimum @xmath214 and @xmath134 .
oscillating @xmath1 and @xmath3 fields behave like ordinary matter satisfying the equation of state @xmath215 @xcite .
the universe becomes isotropic and matter dominated .
the numerical solutions for the stabilizing dilaton potential ( [ eq = v1 ] ) with @xmath211 are plotted in figure [ fig8 ] .
one can see the oscillation of @xmath1 and @xmath101 , and that both @xmath216 and @xmath217 approach to @xmath218 , indicating the matter domination .
damping of @xmath1 and @xmath101 oscillations is due to the expansion of the universe .
( 500,430)(0,0 ) ( 0,240 ) for @xmath211 .
the minimum of the potential is at @xmath219 with @xmath220 .
( b ) , ( c ) and ( d ) show the numerical solutions of @xmath221 , @xmath222 and @xmath223 , respectively , for @xmath211 and initial conditions @xmath224 , @xmath193 and @xmath194 .
@xmath221 , @xmath222 , and @xmath225 approach @xmath226 , @xmath170 , and @xmath227 , respectively.,title="fig:",height=245 ] ( 248,240 ) for @xmath211 .
the minimum of the potential is at @xmath219 with @xmath220 .
( b ) , ( c ) and ( d ) show the numerical solutions of @xmath221 , @xmath222 and @xmath223 , respectively , for @xmath211 and initial conditions @xmath224 , @xmath193 and @xmath194 .
@xmath221 , @xmath222 , and @xmath225 approach @xmath226 , @xmath170 , and @xmath227 , respectively.,title="fig:",height=245 ] ( 2 , 0 ) for @xmath211 .
the minimum of the potential is at @xmath219 with @xmath220 .
( b ) , ( c ) and ( d ) show the numerical solutions of @xmath221 , @xmath222 and @xmath223 , respectively , for @xmath211 and initial conditions @xmath224 , @xmath193 and @xmath194 .
@xmath221 , @xmath222 , and @xmath225 approach @xmath226 , @xmath170 , and @xmath227 , respectively.,title="fig:",height=245 ] ( 240 , 0 ) for @xmath211 .
the minimum of the potential is at @xmath219 with @xmath220 .
( b ) , ( c ) and ( d ) show the numerical solutions of @xmath221 , @xmath222 and @xmath223 , respectively , for @xmath211 and initial conditions @xmath224 , @xmath193 and @xmath194 .
@xmath221 , @xmath222 , and @xmath225 approach @xmath226 , @xmath170 , and @xmath227 , respectively.,title="fig:",height=245 ] ( 120,435)(a ) ( 360,435)(b ) ( 120,195)(c ) ( 360,195)(d )
we investigated cosmology of a four - dimensional low energy effective theory arising from the ns - ns sector of string theory with a d - brane which contains the dynamical degrees of freedom such as the gravity , the dilaton , and the antisymmetric tensor field of second rank , coupling to the gauge field strength living on the brane .
the dynamics of the system crucially depends on the curvature @xmath4 and the brane tension @xmath5 through which the dilaton obtains a potential of the form ; @xmath228 . here , the latter becomes the effective mass of the antisymmetric tensor field ( b - matter ) . in terms of the homogeneous solution in flat spacetime
, we first showed how the dilaton @xmath140 and the non - vanishing magnetic component of the tensor field @xmath141 , in one direction evolve in time . for positive @xmath4
, one finds that the dilaton runs away to negative infinity , and @xmath141 reaches a constant value at late time .
when @xmath4 is negative , one can see that the dilaton is stabilized at a finite value and then the dilaton ( and also @xmath141 ) shows an oscillatory behavior around that value , or runs away again depending on the choice of the initial conditions .
such a dilaton stabilization , however , produces a negative effective cosmological constant , and thus leads to a collapsing universe in string cosmology .
when @xmath141 is turned on , the universe undergoes an anisotropic expansion described by the bianchi type - i cosmology .
we found the attractor solutions showing the overall features of general solutions and confirmed it through numerical analysis .
the dilaton @xmath140 decreases and settles to a logarithmic decrease in time to negative infinity .
when the brane tension term dominates , the anisotropy is sustained . if there is a positive curvature term , it dominates finally over the brane tension term as the dilaton rolls down to negative infinity .
then the expansion of the universe turns to be isotropic and linear in time .
accordingly , @xmath141 decreases inversely proportional to time .
there have been various proposals to stabilize the dilaton . in order to study the dynamics of the b - matter and the stabilized dilaton system
, we adopted an option of generating a dilaton mass term of the form @xmath229 .
then the potential for the dilaton has a global minimum and the cosmological constant of our universe can be fine - tuned to a desired value with negative @xmath4 .
while the dilaton evolved to a stabilized value , @xmath3 shows an oscillatory matter - like behavior , and the universe expands as in the usual matter - dominated era recovering the isotropy .
the obtained result is consistent with that of ref .
this work is the result of research activities ( astrophysical research center for the structure and evolution of the cosmos ( arcsec ) ) and was supported by grant no .
r01 - 2006 - 000 - 10965 - 0 from the basic research program of the korea science @xmath230 engineering foundation ( i.c .
@xmath230 y.k . ) , by the science research center program of the korea science and engineering foundation through the center for quantum spacetime(cquest ) of sogang university with grant number r112005021 ( h.b.k ) .
d. s. goldwirth and m. j. perry , `` string dominated cosmology , '' phys .
d * 49 * , 5019 ( 1994 ) [ arxiv : hep - th/9308023 ] ; k. behrndt and s. forste , `` string kaluza - klein cosmology , '' nucl .
b * 430 * , 441 ( 1994 ) [ arxiv : hep - th/9403179 ] ; e. j. copeland , a. lahiri and d. wands , `` low - energy effective string cosmology , '' phys .
d * 50 * , 4868 ( 1994 ) [ arxiv : hep - th/9406216 ] .
m. kalb and p. ramond , `` classical direct interstring action , '' phys .
d * 9 * , 2273 ( 1974 ) .
e. j. copeland , a. lahiri and d. wands , `` string cosmology with a time dependent antisymmetric tensor potential , '' phys .
d * 51 * , 1569 ( 1995 ) [ arxiv : hep - th/9410136 ] .
e. witten , `` bound states of strings and p - branes , '' nucl .
phys.b * 460 * , 335 ( 1996 ) [ arxiv : hep - th/9510135 ] . e. cremmer and j. scherk , `` spontaneous dynamical breaking of gauge symmetry in dual models , '' nucl .
b * 72 * , 117 ( 1974 ) .
for example , see chapter 3 , 8 and 13 of _ string theory _ " by j. polchinski , cambridge university press ( 1998 ) .
e. j. chun , h. b. kim and y. kim , `` cosmology of antisymmetric tensor field in d - brane universe , '' jhep * 0503 * , 036 ( 2005 ) [ arxiv : hep - ph/0502051 ] . c. kim , h. b. kim , y. kim , o. k. kwon and c. o. lee , `` unstable d - branes and string cosmology , '' j. korean phys .
* 45 * , s181 ( 2004 ) [ arxiv : hep - th/0404242 ] .
h. b. kim , `` dilaton stabilization and inflation in the d - brane world , '' mod .
lett . a * 21 * , 363 ( 2006 ) [ arxiv : hep - th/0505128 ] . s. b. giddings , s. kachru and j. polchinski , `` hierarchies from fluxes in string compactifications , '' phys . rev .
d * 66 * , 106006 ( 2002 ) [ arxiv : hep - th/0105097 ] .
s. kachru , r. kallosh , a. linde and s. p. trivedi , `` de sitter vacua in string theory , '' phys .
d * 68 * , 046005 ( 2003 ) [ arxiv : hep - th/0301240 ] .
l. randall and r. sundrum , `` a large mass hierarchy from a small extra dimension , '' phys .
lett . * 83 * , 3370 ( 1999 ) [ arxiv : hep - ph/9905221 ] .
e. j. copeland , a. r. liddle and d. wands , `` exponential potentials and cosmological scaling solutions , '' phys .
d * 57 * , 4686 ( 1998 ) [ arxiv : gr - qc/9711068 ] .
e. j. copeland , m. sami and s. tsujikawa , `` dynamics of dark energy , '' arxiv : hep - th/0603057 .
a. lukas , b. a. ovrut and d. waldram , `` stabilizing dilaton and moduli vacua in string and m - theory cosmology , '' nucl . phys .
b * 509 * , 169 ( 1998 ) [ arxiv : hep - th/9611204 ] . | we analyze homogeneous anisotropic cosmology driven by the dilaton and the self - interacting `` massive '' antisymmetric tensor field which are indispensable bosonic degrees with the graviton in the ns - ns sector of string theories with d - branes .
we found the attractor solutions for this system , which show the overall features of general solutions , and confirmed it through numerical analysis .
the dilaton possesses the potential due to the presence of the d - brane and the curvature of extra dimensions . in the presence of the non - vanishing antisymmetric tensor field ,
the homogeneous universe expands anisotropically while the d - brane term dominates .
the isotropy is recovered as the dilaton rolls down and the curvature term dominates . with the stabilizing potential for the dilaton
, the isotropy can also be recovered . |
it is well known that the content of a relativistic field theory ( let us take qcd as an example ) may be expressed in terms of functional averages of operators of the type @xmath0 \
> { \cal o}[\overline \psi , \psi , a ] \
> e^{-s_{\rm qcd}[\overline \psi , \psi , a]}\;\;\;. \label{eq : aws path integral}\ ] ] even though it is not possible to perform these integrals exactly , we have learnt an awful lot over the years by studying eq .
( [ eq : aws path integral ] ) in various approximations . for example
, if there is a small parameter ( e.g. @xmath1 , @xmath2 ) one may develop well - ordered expansions of eq .
( [ eq : aws path integral ] ) . alternatively one can use stationary phase methods to approximate the integral relevant , for example , if one is interested in elucidating the importance of classical configurations .
also , these methods are used in order to gain understanding of the behaviour of very high orders of perturbation theory ( hopt ) .
one may study the symmetries of the theory ( e.g. chiral perturbation theory ) in order to relate observables , or one can study its equations of motion ( e.g. dyson - schwinger equation ( dse ) studies ) . finally , much progress has been made in recent years by actually evaluating eq .
( [ eq : aws path integral ] ) directly by discretizing it , as is done in lattice qcd .
we have reported elsewhere on yet another technique , the so - called `` worldline variational approach '' , in which the path integral is approximated , in a rigorous and systematically correctable way , via a variational principle .
this approach was initially applied to relativistic field theory in a scalar model @xcite and more recently to quenched qed @xcite .
the common property that qed shares with qcd is that they are both renormalizable gauge field theories , in contrast to the scalar model which is super - renormalizable .
of course qcd , even after quenching ( i.e. neglecting pair creation ) , differs from quenched qed by the fact that it is asymptotically free whereas quenched qed has a constant coupling .
however , many of the above - mentioned approximative schemes for dealing with eq .
( [ eq : aws path integral ] ) should be applicable within either theory .
quenched qed therefore serves , and is often used , as a test - ground for these methods . in this talk
we compare results for a particular quantity characterising quenched qed , namely its anomalous mass dimension , obtained via a ) the variational approach , b ) via perturbation theory , c ) via dyson - schwinger equation studies and d ) via hopt .
surprisingly , we find that these results are not entirely consistent with each other , which indicates that our present understanding of these methods may be incomplete . in ref .
@xcite we derived from the variational approach an analytic , implicit , result for the anomalous mass dimension @xmath3 of quenched , dimensionally regularized , qed within the ms scheme : @xmath4 this is a remarkable result . at small couplings it should
( approximately ) agree with the known result of perturbation theory has been calculated to fourth order in @xmath5 within su(n ) @xcite ; by suitable choice of casimirs ( i.e. @xmath6 , @xmath7 and , for a quenched theory , @xmath8 ) the result for u(1 ) may be extracted from this .
recently , broadhurst @xcite has also calculated @xmath3 to fourth order directly within qed ] . indeed , as may be seen in fig .
[ fig : aws mass dim ] ( taken from ref .
@xcite ) , the variational result agrees with perturbation theory to within @xmath9 20 % for couplings where 4@xmath10 order perturbation theory appears applicable , i.e. for @xmath5 less than about 1 .
we have also plotted an estimate of perturbation theory up to 5@xmath10 order term of the perturbative expansion order estimate has been derived using the pad approximation methods of ref .
@xcite ] .
it is interesting to note that for a value of @xmath5 around @xmath11 the various orders of perturbation theory appear to ` fan out ' .
it is tempting to speculate that this effect arises because of a finite radius of convergence of the perturbation expansion .
this would be unexpected : the general wisdom from hopt studies is that , as a first rough estimate , the @xmath12 order contribution to a perturbative expansion should scale roughly like the number of diagrams at that order . for the quenched qed propagator , the number of diagrams at order @xmath13 is given by @xmath14 and hence the radius of convergence (= @xmath15 ) of the perturbative expansion vanishes .
more sophisticated analyses@xcite confirm this general result ( although , apparently , the high order behaviour of @xmath16 does not itself appear to ever have been calculated explicitly within quenched qed ) . the final curve plotted in fig .
[ fig : aws mass dim ] is the prediction of a dyson - schwinger equation calculation within the rainbow approximation
. this was obtained , for the dimensionally regularized theory within the ms scheme , in ref .
@xcite and was found to be identical with the well known result for @xmath16 in the theory regularized with a hard cut - off , i.e. @xmath17 . at small couplings this agrees with perturbation theory but then diverges from the latter in a region where ( 4@xmath18 order ) perturbation theory still appears to be applicable . above @xmath19 , @xmath20 becomes complex , which is also the value @xmath21 of the coupling at which the famous chiral symmetry breaking of quenched qed occurs . note also that the perturbative expansion of @xmath20 has a finite radius of convergence given by @xmath22 .
it is , however , not a great surprise that @xmath23 is finite in this calculation as the rainbow approximation contains exactly one diagram at each order and hence does not reproduce the factorial growth ( eq . [ eq : aws factorial growth ] ) in the number of diagrams .
nevertheless , the coincidence of @xmath23 and @xmath24 is interesting and one wonders if it will persist in dyson - schwinger calculations when going beyond the rainbow approximation . also ,
if in the exact theory @xmath25 actually vanishes one wonders what will happen to @xmath21 in that case .
the variational result plotted in fig .
[ fig : aws mass dim ] appears to show no sign of chiral symmetry breaking .
indeed it is easy to derive that at large couplings @xmath26 .
however , it turns out that a perturbative expansion of @xmath27 also has a finite radius of convergence , not too different from the one mentioned above .
we shall discuss this , as well as the large order behaviour of the perturbative expansion of @xmath27 , below .
it is important to note that in this case a finite radius of convergence is not a trivial result of the approximation , as it was for rainbow qed : it can be shown @xcite that each diagram , at any order in perturbation theory , is represented in some approximate way within the variational approximation ; i.e. a perturbative expansion of @xmath27 would receive contributions from @xmath28 diagrams at order @xmath13 .
equation ( [ eq : aws gamvar ] ) is an implicit equation for @xmath27 which can easily be solved numerically for arbitrary @xmath5 .
importantly it can also be solved for complex @xmath5 , so the determination of @xmath25 amounts to a search for nontrivial analytic structure in the complex @xmath5 plane . in fig .
[ fig : aws branch cut ] we show a plot of the ( negative ) real part of @xmath16 as a function of the ( complex ) coupling .
branch cuts , limiting the region of convergence of the perturbation expansion of @xmath27 , are clearly visible for negative @xmath29 .
note that , as opposed to the dyson - schwinger equation result @xmath20 , there are no cuts on the real axis ( also for positive @xmath29 , which is not shown in fig .
[ fig : aws branch cut ] ) .
hence , in fig . [ fig : aws mass dim ] , @xmath27 is real for all @xmath5 while @xmath20 terminates at @xmath25 @xcite .
( 70,70)(15,00 ) ( -60,5)@xmath29 ( -8,20)@xmath30 ( -103,35)-@xmath31 the position of the cuts in fig .
[ fig : aws branch cut ] can be obtained straightforwardly by searching for the value of @xmath16 at which eq .
( [ eq : aws gamvar ] ) has two distinct solutions infinitesimally close to each other ; i.e. the value of @xmath27 at which the derivative with respect to @xmath27 of eq .
( [ eq : aws gamvar ] ) vanishes .
one obtains that the branchpoints are located at @xmath32 , which yields a radius of convergence for the perturbation expansion of @xmath27 of @xmath33 .
the radius of convergence of the perturbative expansion of the anomalous mass dimension can also be obtained , of course , by directly examining the behaviour of its expansion coefficients at large orders , i.e. if we expand @xmath34 then radius of convergence is given by @xmath35 for @xmath13 up to @xmath36 , it is possible to obtain these expansion coefficients by direct substitution of eq .
( [ eq : aws perturbative expansion ] ) into eq .
( [ eq : aws gamvar ] ) .
the results are tabulated in table [ table : aws cn ] , where it is clearly seen that although @xmath37 does perhaps tend to a finite limit , the rate of convergence is rather slow ( and non - uniform ) .
ccccccc n & 5 & 10 & 15 & 20 & 25 & 30 + + @xmath38 & 2.00 & 1.31 & 1.16 & 1.11 & 1.02 & 1.16 + in order to find the @xmath39 s for higher values of @xmath13 , it proves to be advantageous to convert eq .
( [ eq : aws gamvar ] ) into a differential equation in order to eliminate the cotangent
. we obtain @xmath40 if we now substitute eq .
( [ eq : aws perturbative expansion ] ) for @xmath27 , and @xmath41 for @xmath42 ( hence @xmath43 and @xmath44 ) we may solve for the coefficients @xmath39 and @xmath45 in an iterative manner .
the results are shown in fig . [
fig : aws cn ] .
( 50,50)(-65,00 ) the slow rate of convergence , as well as its non - uniformity , is clearly visible in this figure .
indeed , numerically one finds that for large @xmath13 the points in fig .
[ fig : aws cn ] are fitted exceedingly well by the functional form @xmath46\;\;\;,\ ] ] with @xmath47 , @xmath48 and @xmath49 .
these values of @xmath39 may be compared to the equivalent expansion coefficients obtained from the dyson - schwinger equation result : @xmath50 in fact , the variational result and the rainbow dyson - schwinger result for the @xmath39 s are surprisingly similar , not only in their functional form but even in the numerical coefficients .
the main difference is the occurance of the sine function in the former .
it is because of this sine that the branchcut , which for @xmath20 lies on the real axis , has moved into the complex plane for @xmath27 .
in this contribution we have compared predictions for the anomalous mass dimension of quenched qed obtained through the use of a variety of techniques : the worldline variational approach , perturbation theory up to @xmath51 $ ] ( as well as a pad estimate for the @xmath52 $ ] term ) , rainbow dyson - schwinger equation studies and general expectations from studies of high order perturbation theory .
both the rainbow dse s and the variational approach yield a cut in @xmath16 as a function of the coupling . in the former
this cut is on the real axis , and is associated with chiral symmetry breaking , while in the latter it has moved ( a long way ) off the real axis , so that in that calculation there is no obvious sign of chiral symmetry breaking . in either case , the cuts necessitate a finite radius of convergence for the perturbative expansion of @xmath3 , and one can argue that circumstantial evidence for this may be seen in the perturbative result as well .
general expectations from hopt , on the other hand , suggest that the radius of convergence should be zero .
clearly , as quenched qed is the prototype gauge theory for the investigation of chiral symmetry breaking , efforts should be made to clarify these disagreements .
it may even be the case that the eventual resolution of these issues will teach us something about the limitations of one or more of these approximate techniques for dealing with nonperturbative gauge field theories .
rosenfelder , r. and schreiber , a. w. , ; _ ibid _ 33543365 ( 1996 ) ; schreiber , a. w. , rosenfelder , r. and alexandrou , c. , ; schreiber , a. w. and rosenfelder , r. , ; alexandrou , c. , rosenfelder , r. and schreiber , a. w. , .
alexandrou , c. , rosenfelder , r. and schreiber , a. w. , submitted to , hep - th/0003253 ; see also schreiber , a. w. , rosenfelder , r. and alexandrou , c. , contribution to appear in `` proceedings of the workshop on light - cone qcd and nonperturbative hadron physics '' , adelaide , dec .
1999 , adp-00 - 16/t400 . for a review of hopt see , for example ,
bogomolnyi , e. , fateev , v. a. and lipatov , l. n. , ; a useful collection of papers on the subject may be found in `` large - order behaviour of perturbation theory '' , eds .
j.c . le guillou and j. zinn - justin , north - holland ( 1990 ) ; an explicit calculation of the electron s anomalous magnetic moment calculated within quenched qed has been carried out in itzykson , c. , parisi , g. and zuber , j - b . , ( paper 43 of the above volume ) . | in order to understand qcd at the energies relevant to hadronic physics one requires analytical methods for dealing with relativistic gauge field theories at large couplings . strongly coupled quenched qed provides an ideal laboratory for the development of such techniques , in particular as many calculations suggest that like qcd this theory has a phase with broken chiral symmetry . in this talk
we report on a nonperturbative variational calculation of the electron propagator within quenched qed and compare results to those obtained in other approaches .
we find surprising differences among these results .
adp-00 - 20/t403 |
the striking feature of the lightcurves of known gamma - ray pulsars are relatively long duty cycles as well as phase shifts in comparison to the radio pulses ( thompson et al . 1999 , thompson 2001 , kanbach 2002 ) .
the lightcurve shapes fall into three categories .
the three brightest gamma - ray pulsars - the crab pulsar , vela and geminga , along with b1951 + 32 exhibit two well defined , sharp peaks separated in phase by @xmath0 , and connected by the interpeak bridge of a considerable level . b1706 - 44 and b1055 - 52 show two peaks separated by about 0.2 in phase , whereas b1509 - 58 exhibits a broad single pulse .
similar properties are present in the x - ray domain of the high - energy emission . moreover , in the case of the crab pulsar its optical pulsed emission is usually considered jointly with the high - energy emission : the phase - averaged spectra in gamma and x - rays connect smoothly to optical , and the pulses have similar shapes and phases .
this suggests that gammas , x - rays and optical light may come from the same regions . these properties , along with the spectral properties which are not the subject of this paper , prompted substantial refinements of the physical models of pulsar high - energy activity , both outer - gap models ( cheng et al .
1986 , romani & yadigaroglu 1995 , zhang & cheng 1997 ) as well as polar - cap models ( sturrock 1971 ; ruderman & sutherland 1975 ; daugherty & harding 1982 , sturner et al .
1995 ) , ( see rudak , dyks & bulik 2002 for recent critical review ) . despite reiterated arguments in favour of polar - gap models ( e.g. baring 2001 ) and outer - gap models ( e.g. yadigaroglu & romani 1995 ) , both classes of models still suffer from serious problems . according to the polar cap model , the characteristic double - peak lightcurve forms when the line of sight intersects the polar - cap beam where the highest - energy emission originates : upon entering the beam a leading peak is produced , followed by a bridge emission due to the inner parts of the beam , when the line of sight exits the beam the trailing peak forms . because of long duty cycles in the high - energy domain on one hand , and narrow opening angles for gamma - ray emission on the other hand , polar cap models have to rely on nearly aligned rotators , where inclination ( the angle @xmath1 ) of the magnetic axis ( the magnetic moment @xmath2 ) to the spin axis ( the angular velocity @xmath3 ) is comparable to the angular extent of the polar cap ( daugherty & harding 1994 ) . the outer - gap model , on the contrary , prefers highly inclined rotators ( chiang & romani 1992 ; romani & yadigaroglu 1995 ; cheng et al .
however , the model is unable ( without additional postulates ) to account for the presence of outer wings in the double - peak lightcurves .
more importantly , the model in its present form ( e.g. zhang & cheng 2002 ) is unable to account for a substantial level of the off - pulse emission in the crab pulsar . also on theoretical grounds
the existence of outer gaps in the present - day ` vacuum ' approach ( i.e. with the gap extending between the null surface and the light cylinder ) has been questioned .
the outer gap remains anchored to the conventional null surface , provided that no current is injected at the boundaries of the accelerator ( no pairs are created ) ; otherwise the gap extention becomes very sensitive to the details of pair creation ( hirotani & shibata 2001 , hirotani et al .
2003 ) .
these problems motivated us to propose a new picture of the origin of high - energy radiation within the pulsar magnetospheres - in regions confined to the surface of last open magnetic field lines ( similarly to thin outer - gap accelerators ) but extending between the polar cap and the light cylinder .
the most important consequence of such extended accelerators , as far as the lightcurves are concerned , is a caustic nature of the high - energy peaks : special relativity effects ( aberration of photon emission directions and time of flight delays due to the finite speed of light @xmath4 ) cause that photons emitted at different altitudes within some regions of the magnetosphere are piled up at the same phase of a pulse ( morini 1983 ; romani & yadigaroglu 1995 ) .
the term ` high - energy radiation ' used throughout the paper refers to nonthermal radiation in the energy domain of gamma - rays and hard x - rays ( i.e. above several kev ) . the soft x - ray radiation ( @xmath5 kev ) is not included into our considerations because it is heavily affected by thermal emission from the neutron star surface in some objects , like geminga which exhibits a complex pattern of soft x - ray pulses ( e.g. fig.4 in jackson et al .
2002 ) . for reasons mentioned at the beginning of this section , in the case of the crab pulsar our definition includes also the optical band .
the paper is arranged in the following way : in section 2 we introduce the two - pole caustic model for the high - energy lightcurves of pulsars and present the results of numerical calculations .
section 3 contains discussion of the generic features of the model as well as comparison with the properties of other models .
special relativity ( sr ) effects which affect pulsar lightcurves include the aberration of photon emission directions and the time of flight delays caused by the finite speed of light @xmath4 .
morini ( 1983 ) was the first to prove that these effects were able to produce prominent peaks in pulsar lightcurves .
he obtained the peaks of caustic origin in his version of the polar cap model , which included photon emission from high altitudes , where the sr effects are important . unlike in the case of the morini s model , in the standard polar cap model the strongest gamma - ray emission takes place very close to the star surface , where the caustic effects are not important and ,
therefore , pulsar lightcurves are mostly determined by the altitudinal extent of the accelerator ( polar gap ) and by the emissivity profile along the magnetic field lines ( eg . daugherty & harding 1996 ; dyks & rudak 2000 ) . in the recent version of the outer gap model ( chiang & romani 1992 ; romani & yadigaroglu 1995 ; yadigaroglu 1997 ; cheng et al .
2000 ) the peaks in the lightcurves are purely due to the caustic effects the fact first emphasized by romani & yadigaroglu ( 1995 ) . however , the altitudinal extent of accelerator ( limited by the position of the inner boundary of the outer gap ) is crucial in limiting the possible shapes of lightcurves within this model .
interestingly , yadigaroglu ( 1997 ) also considered photon emission along the entire length of all last open magnetic field lines , relaxing thus the outer - gap extent limit ( see the bottom panel in fig .
3.1 of his thesis ) .
however , he did nt pursue this possibility in any further details .
hereby we propose that the observed high - energy emission from known gamma - ray pulsars originates from the regions considered already by yadigaroglu ( 1997 ) .
let us assume that the gap region ( the region where particle acceleration is taking place as well as high - energy photons originate ) posesses the following properties : + - the gap region extends from the polar cap to the light cylinder , + - the gap is thin and confined to the surface of last open magnetic - field lines , + - photon emissivity is uniform within the gap region .
+ fig.1 shows schematically the location and the extension of the proposed gap ( for the sake of comparison the location of the outer gap is shown as well ) .
the resulting lightcurves are dominated by strong peaks ( either double or single ) of caustic origin . in the outer - gap model ,
the inner boundary of the gap is located at an intersection of the null - charge surface with the last closed field lines .
its radial distance @xmath6 is , therefore , a function of azimuthal angle @xmath7 , with a minimum value @xmath8 in the @xmath9-plane .
for highly inclined rotators @xmath8 becomes a small fraction of the light cylinder radius @xmath10 : @xmath11 ^ 2 $ ] ( halpern & ruderman 1993 ) ; for the inclination angle @xmath12 the ratio is @xmath13 .
however , for azimuthal directions departing the @xmath9-plane , the inner radius @xmath6 tends to @xmath14 .
the radiation region extends out to the light cylinder and the radiation escaping the magnetospheric region comes from particles moving outwards along the magnetic lines .
+ in our model we assume , however , that the actual gap extends to the polar cap , i.e. @xmath15 for all azimuthal angles @xmath7 .
this assumption is essential for the expected performance of the model since it implies that the observer can detect radiation originating from _ both _ magnetic hemispheres . as in the outer - gap model of chiang & romani ( 1992 ) , and
cheng et al . ( 2000 ) we assumed that the photon emissivity is uniform everywhere in the proposed gap region . for simplicity , we consider a rigidly rotating static - like magnetic dipole .
departures of the retarded field lines from the static case are of the order of @xmath16 and they are insignificant since prominent features in modelled lightcurves ( peaks ) which we discuss below arise due to radiation at radial distances @xmath17 . also , rotationally - driven currents can be neglected : longitudinal currents suspected to flow within the open field line region can not modify @xmath18 by a factor exceeding @xmath19 whereas toroidal currents due to plasma corotation change @xmath18 barely by @xmath16 ( @xmath20 is the local corotation velocity in the speed of light units ; @xmath21 ) .
we considered radiating particles moving from the outer rim of the polar cap toward the light cylinder .
the photons were emitted tangentially to local magnetic field lines in the corotation frame , and then they were followed crossing the magnetosphere with no magnetic attenuation .
our treatment of rotational effects in the numerical code used for the calculations is the same as in yadigaroglu ( 1997 ) and in dyks & rudak ( 2002 ) .
specifically , we used the following standard aberration formula to transform the unit vector of photon propagation direction @xmath22 from the corotating frame to its value @xmath23 in the inertial observer frame : @xmath24 where @xmath25 .
the above formula results directly from the general lorentz transformation . by replacing @xmath26 and @xmath27 in the formula with @xmath28 and @xmath29 ( respectively ) , one obtains a general lorentz transformation for a velocity vector @xmath30 .
+ photon travel delays were taken into account by adding @xmath31 to the azimuth @xmath32 of photon emission direction @xmath23 ( @xmath33 is the radial position of emission point , and @xmath23 is the unit vector of photon propagation direction after aberration effect is included ) .
the result , taken with a minus sign , is a phase of detection @xmath34 : @xmath35 our numerical code performs a raytracing followed by a numerical integration of the observed pulse profile .
all results described below are just due to the dipolar shape of the magnetic field , the aberration effect ( eq . 1 ) , as well as the light travel time delays ( eq . 2 ) .
the results are presented in fig.[fig2 ] in the form of photon mapping onto the @xmath36-plane , with accompanying lightcurves for five viewing angles @xmath37 ( the angle between the spin axis and the line of sight ) .
inclination angle @xmath12 , and spin period @xmath38 s were assumed for the rotator . for each magnetic pole two caustics form in the @xmath36-plane :
1 ) the dominant caustic , and 2 ) the subdominant caustic . + the dominant caustic ( easy to identify in the photon mapping ) is associated with the trailing part of the emission region with respect to its magnetic pole .
the subdominant caustic ( much weaker ) is associated with the leading part of the emission region .
characteristic features in the lightcurves due to caustic crossing for large values of @xmath37 are marked with capital letters from a to d. the dominant - caustics crossing yields two prominent peaks d and b , in the lightcurve .
the peaks consist of photons emitted over a very wide range of altitudes ( e.g. for @xmath39 almost all altitudes between the star surface and the light cylinder contribute to the peaks ; see also fig . 2 in morini 1983 ) .
the subdominant - caustics crossing yields the features a and c , which actually contribute to the trailing wings of the peaks d and b , respectively . features a and c consist of photons emitted very close to the light cylinder ( cf .
fig . 9 in cheng
this result is in clear contrast with the results obtained for the conventional outer - gap model ( see fig.[fig3 ] ) , where the leading peak ( a ) is due to the subdominant caustic , and the trailing peak ( b ) is due to the dominant caustic . unlike in the outer gap model , in our model each of the two prominent peaks in the resulting lightcurve
is associated with a different magnetic pole .
therefore , we propose the name two - pole caustic model " .
the position of the last closed magnetic field lines and the magnitude of special relativity effects are governed by the proximity of the light cylinder and , therefore , the radiation pattern as well as the lightcurves shown in fig . 2 do not depend on the rotation period @xmath40 but solely on the inclination angle @xmath1 ( with one exception : the size of blank spots corresponding to polar caps does depend on @xmath40 ) .
therefore , the two - pole caustic model is relevant for pulsars with any rotation period .
we find that noticeable peaks of caustic origin appear in the lightcurves practically for any inclination of magnetic dipole @xmath41 and for any viewing angle @xmath42 . for small inclinations ( @xmath43 )
the peaks are broader and it is more probable to observe single - peaked lightcurves ; such lightcurves form for a wide range of viewing angles for which the coventional outer gap model predicts no emission . in the glast era
, the detectability of moderately inclined pulsars " viewed far from the equatorial plane will serve as a clear discriminator between the two - pole caustic model and the traditional outer gap model .
[ fig4 ] presents a comparison of a lightcurve calculated within the two - pole caustic model for the vela pulsar with a lightcurve obtained for this pulsar with egret ( kanbach 1999 ) .
the model lightcurve was calculated for electrons distributed evenly along the polar cap rim ; the density profile across the rim was assumed to be the gaussian function @xmath44 , symmetrical about @xmath45 , with @xmath46 ( @xmath47 is the magnetic colatitude of magnetic field lines footprints at the star surface , and @xmath48 is the magnetic colatitude of the rim ) .
photon emission was followed up to @xmath49 .
the shape of the egret lightcurve is very well reproduced by the two - pole caustic model : the leading peak is narrower than the trailing peak , which connects smoothly with the bridge emission ; the leading peak seems to present a separate entity it does not connect smoothly with the bridge , and is followed by a characteristic ` interpulse bump ' .
these features result generically from the two - pole caustic model , since it predicts that the trailing peak , the bridge emission , and the ` interpulse bump ' arise from sampling a single , continuous radiation pattern from one magnetic pole ( cf .
[ fig2]a ) .
the leading peak and the offpulse emission are produced by sampling an emission pattern from the opposite pole .
the ` interpulse bump ' predicted by the two - pole caustic model is not observed in the crab pulsar .
this might be caused by a decline in the photon emissivity above @xmath50 .
single peaked gamma - ray lightcurves as the one observed for b1509@xmath5158 ( kuiper et al .
1999 ) can be reproduced in the two - pole caustic model for small viewing angles ( fig .
[ fig2]b ) , however , the predicted phase lag between the gamma - ray and the radio peak ( @xmath52 ) is smaller than the one observed for b1509@xmath5158 ( @xmath53 ) .
moreover , double - peak lightcurves with small separation between the peaks in b1706@xmath5144 ( thompson et al .
1996 ) , and b1055@xmath5152 ( thompson et al .
1999 ) can not be interpreted in the same way as the lightcurves with wide separation of the peaks .
these difficulties forced chiang & romani ( 1992 ) ( and probably yadigaroglu 1997 ) to abandon the geometry of the two - pole caustic model ( r. w. romani , private communication ) .
we agree with r. w. romani that these particular lightcurves should be interpreted in terms of the outer gap caustics a and b ( figs .
[ fig2 ] and [ fig3 ] ) .
this interpretation can be accommodated by the two - pole caustic model only when the outer gap part of the lightcurves ( between a and b in fig .
[ fig2 ] ) dominates over the leading peak formed by the trailing caustic d. as can be noticed in fig .
[ fig2 ] , the intensity of the outer gap part of lightcurves ( between a and b ) increases with respect to the intensity of the peak d , as the viewing angle @xmath37 approaches the value for which the line of sight barely skims the outer gap part of the radiation pattern . for
some viewing gometries the intensity of the outer gap part ( a b ) may exceed by a few times the intensity of the peak d. fig .
[ fig5 ] shows an example of such a lightcurve .
given the low intensity and large width of the peak d , it may stay unresolved in the low statistics data of b1509@xmath5158 , b1706@xmath5144 , and b1055@xmath5152 .
table 1 summarizes major similarities and differences between our model and other models . for these purposes we choose the model of morini 1983 ( it was the first model where caustic effects were noticed ) and the model of smith et al .
1988 , along with the polar - cap model and the outer - gap model .
we introduced a two - pole caustic model for the high - energy lightcurves of pulsars ( for the crab pulsar in particular ) .
the effects of aberration and light travel delays , as well as the geometry of the last closed magnetic field lines are essential for forming the lightcurves of a caustic nature .
the generic features in the lightcurves provided by the two - pole caustic model are consistent with the observed characteristics of pulsar lightcurves : + 1 .
the most natural ( in terms of probability ) shape consists of two peaks ( separated by 0.4 to 0.5 in phase for large viewing angles ) , + 2 .
the peaks posess well developed wings , + 3 .
there is a bridge ( inter - peak ) emission component , + 4 .
there is a non - vanishing off - pulse emission level , + 5 .
the radio pulse ( or pulse precursor , in the case of crab ) comes ahead of the leading peak ( by @xmath52 in phase for large viewing angles ) . features 1 . , 3 . , and 5 . are not a unique property of the two - pole caustic model , they can be easily obtained within the outer - gap model ( compare the lightcurves in figs.[fig2 ] and [ fig3 ] ) .
feature 2 .
may in principle be obtained within the outer - gap model , but no consensus exists among the proponents of that model on the actual physical reason behind this feature ( yadigaroglu 1997 , cheng , ruderman & zhang 2000 ) . in our model
the trailing wings are formed by the subdominant caustics ( a ) and ( c ) which often blend with peaks ( d ) and ( b ) , respectively ( the effect of blending is not shown in fig . 2 since the calculation were truncated at @xmath54 .
) feature 4 .
, however , may play a decisive role in showing the advantage of the two - pole caustic model over the outer - gap model : this particular feature of our model may explain the presence of the significant x - ray flux from the crab pulsar at pulse minimum discovered by tennant et al .
it has not been demonstrated so far how such a feature could be obtained within the outer - gap model ; in particular , it is absent in the x - ray pulse profile calculated for the crab pulsar in the recent model of zhang & cheng 2002 .
an interesting property of the double - peak lightcurve , inherent to the two - pole caustic model , emerges for viewing angles @xmath37 close to @xmath55 : the trailing peak ( together with its wings ) assumes the shape which is roughly similar ( in the sense of translations in the rotation - phase @xmath34 space ) to the shape of the leading peak and its wings ( cf .
fig.[fig2 ] ) .
such effect is not possible in the case of the outer - gap model ( cf .
fig.[fig3 ] ) , nor it is possible in the polar cap model ( see wona et al .
2002 ) ; these two models lead to approximate ` mirror'-symmetry in the double - peak lightcurves . in principle then , this property might also be used to discriminate between the two - pole caustic model and other models . an important testing - ground for any models will be polarization properties of the high - energy radiation . for the time being , good - quality polarization information is available only for optical light from crab ( smith et al . 1988 ) . an argument in favor of the caustic origin of the optical peaks of the crab pulsar is that the degree of polarization drops to minimum values at the phases of both peaks ( cf .
4c of smith et al .
such a drop is justified by virtue of the caustic nature of the peaks : it results from a pile - up of polarized radiation with different polarization angles .
smith et al .
@xcite emphasize , that the behaviour of the polarization as a function of rotation - phase for crab is strikingly similar for both peaks , i.e. the polarization behaviour at the phase of the leading peak repeats at the trailing peak .
the outer gap model is able to reproduce this feature , even though each of the two peaks arises from a very different type of caustic in this model : the leading peak is due to the caustic formed close to the light cylinder at the _ leading _ part of the emission region , whereas the trailing peak is due to the caustic formed within the _ trailing _ part .
romani & yadigaroglu ( 1995 ) consider the ability to reproduce the double sweep in the polarization position angle to be one of major successes of the outer - gap model . in the two - pole caustic model ,
the two peaks arise due to crossing the same type of caustic - the dominant caustic associated with the trailing part of the emission region ( cf .
section 2 ) .
we suspect , therefore , that such a double sweep should even more naturally be produced by the two - pole caustic model . a comprehensive treatment of the polarization properties of high - energy radiation in the two - pole caustic model will be the subject in our future work .
we emphasize that the characteristic form of the high - energy lightcurves of pulsars ( the double - peak structure with bridge emission in high - energy , preceded by the peak in radio ) is an inherent property of a rotating source with a magnetic dipole , with roughly uniform high - energy emissivity along the last open field lines .
two recently proposed models may provide physical grounds for the geometry of the two - pole caustic model : the slot - gap model of arons & scharlemann ( 1979 ) , in the modern version of muslimov & harding ( 2003 ) ; and the model of hirotani et al .
( 2003 ) of an outer gap extended on either side of the null surface due to the currents . when a realistic physical model for the extended gaps becomes available
, calculations of spectral characteristics within our model will be possible .
we acknowledge fruitful discussions with a. harding , k. hirotani and a. muslimov .
we thank j. gil for bringing our attention to the problem of missing wings in the double - peak lightcurves in the outer - gap model and r. w. romani for pointing out some shortcomings of the two - pole caustic model .
comments by the anonymous referee helped us to clarify the paper significantly .
this work was supported by the grant pbz - kbn-054/p03/2001 .
part of the work was performed while jd held a national research council research associateship award at nasa / gsfc .
arons , j. & scharlemann , e. t. , 1979 , , 231 , 854 baring , m. g. 2001 , proc .
tonantzintla workshop , eds .
a. carrami~ nana et al . ,
astrophysics and space science library , volume 267 ( kluwer academic publishers dordrecht ) , p.167 cheng , k. s. , ho , c. , & ruderman , m. 1986 , , 300 , 500 cheng , k. s. , ruderman , m. a. , & zhang , l. 2000 , , 537 , 964 chiang , j. , & romani , r. w. 1992 , , 400 , 629 kuiper , l. , hermsen , w. , krijger , j. m. , bennett , k. , carramiana , a. , et al .
1999 , a&a , 351 , 119 daugherty , j.k .
, & harding , a.k .
1982 , apj , 252 , 337 daugherty , j.k . , & harding , a.k .
1994 , apj , 42 , 325 daugherty , j.k . , &
harding , a.k .
1996 , apj , 458 , 278 dyks , j. , & rudak , b. 2000 , , 319 , 477 dyks , j. , & rudak , b. 2002 , a&a , 393 , 511 halpern , j. p. , & ruderman , m. a. 1993 , , 415 , 286 hirotani , k. , & shibata , s. , 2001 , , 325 , 1228 hirotani , k. , harding , a.k . , & shibata , s. , 2003 , , 591 , 334 jackson , m. s. , halpern , j. p. , gotthelf , e. v. , & mattox , j. r. 2002 , , 578 , 935 kanbach , g. , 1999 , in proc . of the third integral workshop , eds .
a. bazzano , g. g. c. palumbo , & c. winkler , astrophys .
comm . , 38 , 17 kanbach , g. , 2002 , in proc . of the 270 .
we - heraeus seminar on neutron stars , pulsars , and supernova remnants , eds .
w. becker , h. lesch , and j. trmper , mpe report 278 , 91 ( astro - ph/0209021 ) morini , m. 1983 , , 202 , 495 muslimov , a. , & harding , a.k . , 2003 , , 588 , 430 romani , r. w. & yadigaroglu , i .- a .
, 1995 , , 438 , 314 rudak , b. , dyks , j. , & bulik , t. , 2002 , in proc .
of the 270 .
we - heraeus seminar on neutron stars , pulsars , and supernova remnants , eds .
w. becker , h. lesch , and j. trmper , mpe report 278 , 142 ( astro - ph/0206101 ) ruderman , m. a. , & sutherland , p. g. 1975 , , 196 , 51 smith , f. g. , et al .
1988 , , 233 , 305 sturner , s. j. , dermer , c. d. , & michel , f. c. 1995 , , 445 , 736 sturrock , p. a. 1971 , , 164 , 529 tennant , a. p. , et al .
2001 , , 554 , l173 thompson , d. j. , bailes , m. , bertsch , d. l. , esposito , j. a. , fichtel , c. e. , et al .
1996 , , 465 , 385 thompson , d. j. , bailes , m. , bertsch , d. l. , cordes , j. , damico , n. , et al . , 1999 , , 516 , 297 thompson , d. j. , 2001 , gamma - ray pulsars : observations , in high energy gamma - ray astronomy , aip proceedings , eds .
f. a. aharonian and h. j. v " olk , 558 , 103 yadigaroglu , i .- a .
1997 , ph.d .
thesis , stanford university yadigaroglu , i .- a . , & romani , r. w. 1995 , , 449 , 211 wona , a. , dyks , j. , rudak , b. , & bulik , t. 2002 , proc . of the xxviii moriond meeting the gamma - ray universe , eds .
a. goldwurm et al . , 539 ( astro - ph/0205224 ) zhang , l. , & cheng , k. s. 1997 , , 487 , 370 zhang , l. , & cheng , k. s. 2002 , , 569 , 872 & @xmath59 & @xmath60 + caustic origin of the peaks & @xmath51 & @xmath61 @xmath62 & @xmath51 & + & + + -1.7mm@xmath63 & @xmath51 & @xmath51 & + & @xmath51 & + + -1.7mm@xmath64 & + & + & @xmath51 & @xmath51 & + + ( the acceleration region is extended ) & @xmath51 & @xmath51 & @xmath51 & + & + + | we present a new model of high - energy lightcurves from rotation powered pulsars .
the key ingredient of the model is the gap region ( i.e. the region where particle acceleration is taking place and high - energy photons originate ) which satisfies the following assumptions : i ) the gap region extends from each polar cap to the light cylinder ; ii ) the gap is thin and confined to the surface of last open magnetic - field lines ; iii ) photon emissivity is uniform within the gap region .
the model lightcurves are dominated by strong peaks ( either double or single ) of caustic origin . unlike in other pulsar models with caustic effects ,
the double peaks arise due to crossing two caustics , each of which is associated with a different magnetic pole .
the generic features of the lightcurves are consistent with the observed characteristics of pulsar lightcurves : 1 ) the most natural ( in terms of probability ) shape consists of two peaks ( separated by 0.4 to 0.5 in phase for large viewing angles ) ; 2 ) the peaks posess well developed wings ; 3 ) there is a bridge ( inter - peak ) emission component ; 4 ) there is a non - vanishing off - pulse emission level ; 5 ) the radio pulse occurs before the leading high - energy peak .
the model is well suited for four gamma - ray pulsars - crab , vela , geminga and b1951 + 32 - with double - peak lightcurves exhibiting the peak separation of 0.4 to 0.5 in phase .
hereby , we apply the model to the vela pulsar .
moreover , we indicate the limitation of the model in accurate reproducing of the lightcurves with single pulses and narrowly separated ( about 0.2 in phase ) pulse peaks .
we also discuss the optical polarization properties for the crab pulsar in the context of the two - pole caustic model . |
the author is grateful to v.o .
manturov for idea of @xmath0-polynomial for long virtual knots and fruitful consultations . | we construct new invariant polynomial for long virtual knots .
it is a generalization of alexander polynomial . we designate it by @xmath0 meaning an analogy with @xmath0-polynomial for virtual links .
a degree of @xmath0-polynomial estimates a virtual crossing number .
we describe some application of @xmath0-polynomial for the study of minimal long virtual diagrams with respect number of virtual crossings .
virtual knot theory was invented by kauffman around 1996 @xcite .
long virtual knot theory was invented in @xcite by m. goussarov , m. polyak , and o. viro .
@xmath0-polynomial for virtual link was introduced independently by several authors ( see @xcite,@xcite,@xcite,@xcite ) , for the proof of their coincidence , see @xcite .
the idea of two types of classical crossings in a long diagram , which were called @xmath1 ( circle ) and @xmath2 ( star ) , was invented by v.o .
manturov ( see @xcite,@xcite ) . in present paper we called @xmath1 and @xmath2 crossings by _ early overcrossing _ and _ early undercrossing _ respectively . to consider early overcrossings and early undercrossings is the basis idea for a construction of @xmath0-polynomial in the case of long virtual knots . by a _ long virtual knot diagram _ we mean a smooth immersion @xmath3 such that : \1 ) outside some big circle
, we have @xmath4 ; \2 ) each intersection point is double and transverse ; \3 ) each intersection point is endowed with classical ( with a choice for underpass and overpass specified ) or virtual crossing structure .
a _ long virtual knot _ is an equivalence class of long virtual knot diagrams modulo generalized reidemeister moves . by an _
arc _ of a long virtual knot diagram we mean a connected component of the set , obtained from the diagram by deleting all virtual crossings ( at classical crossing the undercrossing pair of edges of the diagram is thought to be disjoint as it is usually illustrated ) .
we say that two arcs @xmath5 belong to the same _ long arc _ if there exists a sequence of arcs @xmath6 and virtual crossings @xmath7 such that for @xmath8 the arcs @xmath9 are incident to @xmath10 from opposite sides . throughout the paper , we mean that initial and final long arcs , @xmath11 and @xmath12 , form united long arc @xmath13 .
let @xmath14 be a long virtual diagram with @xmath15 classical crossings .
hence , there is a natural pairing of all classical crossings and all long arcs : classical crossing @xmath16 and long arc @xmath17 , which emanates from @xmath16 , are paired .
we say that classical crossing @xmath16 is _ early overcrossing _
( _ early undercrossing _ ) if we have an arc passing over ( under ) @xmath16 at first , in the natural order on long virtual diagram ( see also @xcite , p. 139 ) .
an _ incidence coefficient _
t=\mathbb{z}[p , p^{-1},q , q^{-1}]/((p-1)(p - q),(q-1)(p - q))$ ] of classical crossing @xmath16 and arc @xmath19 is defined as a sum of some of three polynomials : @xmath18={\varepsilon}_1 1+{\varepsilon}_2 ( t^{sgn\,v}-1)+{\varepsilon}_3(-t^{sgn\,v})$ ] , where @xmath20 ; @xmath21 if @xmath16 is early overcrossing , @xmath22 if @xmath16 is early undercrossing ; @xmath23 denotes _ local writhe number _ of @xmath16 .
we set @xmath24 arc @xmath19 is emanating from @xmath16 ; @xmath25 @xmath19 is passing over @xmath16 ; @xmath26 @xmath19 is coming into @xmath16 . if @xmath16 and @xmath19 are not incident we set @xmath18=0 $ ] .
let us enumerate all classical crossings of @xmath14 by numbers @xmath27 in arbitrary way and associate with each classical crossings the emanating long arc .
our generalization of alexander polynomial for long virtual knots is defined as determinant of @xmath28-matrix @xmath29 with elements @xmath30s^{deg\,a}\in t[s , s^{-1}]\ ] ] the function @xmath31arcs of d@xmath32 is defined according to the rules : \(1 ) if arc @xmath19 is a first at a long arc , @xmath33 ; \(2 ) if arcs @xmath19 and @xmath34 are neighbour on a long arc , @xmath19 precedes @xmath34 , then @xmath35 , if we pass from the left to the right with respect to the transversal arc , and @xmath36 otherwise .
in the first case we called such virtual crossing _ increasing _
, in the second case
_ decreasing_. it easy to see that polynomial @xmath37 does not depend on a numeration of classical crossings . by analogy with @xcite
we formulate following three theorems .
[ theorem0 ] if virtual diagrams @xmath38 are equivalent then @xmath39 for some integer @xmath40 .
the invariance of @xmath0 for reidemeister moves @xmath41 is evident .
the checking of invariance for @xmath42 and @xmath43 is similar to the case of @xmath0-polynomial for virtual link ( see @xcite,@xcite ) .
there are two types of the first reidemeister move @xmath44 : @xmath45 , if we have early overcrossing , and @xmath46 , if we have early undercrossing .
it easy to calculate that @xmath47 , @xmath48 .
it is convenient to use the laplace theorem ( about determinants ) to check that @xmath49 .
we check equality for @xmath50 pair of @xmath51-minors of matrices @xmath52 and @xmath53 .
two of these pairs give equalities only if we set @xmath54 .
[ theorem1 ] let @xmath55 be the number of virtual crossings on a long virtual diagram @xmath14 .
then @xmath56 . from theorems [ theorem0 ] and [ theorem1 ]
we easily conclude [ cor1 ] if @xmath57 then @xmath14 has minimal virtual crossing number .
for checking of minimality by using corollary [ cor1 ] it is convenient to use [ theorem3 ] the @xmath58-th coefficient of @xmath59 is equal to @xmath60 , where @xmath61 $ ] if @xmath62 s.t .
@xmath63#of increasing virtual crossings on @xmath64 , and @xmath65 otherwise , @xmath66 . example . in * figure
* we draw long virtual diagram @xmath67 which closure is unknot .
arcs @xmath68 , @xmath69 , are marked by thick lines . by theorem [ theorem3 ]
the @xmath58-th coefficient of @xmath59 is equal to @xmath70|_{i , j=1, ... ,n}=$ ] @xmath71 in the ring @xmath72 .
consequently , @xmath67 is minimal by corollary [ cor1 ] . by using our @xmath0-polynomial
we can proof following conjecture in a particular case . here
symbol @xmath73 denotes usual product of long knots .
conjecture .
_ if @xmath14 is a minimal long virtual diagram with respect number of virtual crossings , k is a long classical knot diagram , then @xmath74 is also minimal . _
[ theorem2](the particular case of conjecture ) + if @xmath14 is a minimal long virtual diagram s.t .
@xmath75 is equal to virtual crossing number of @xmath14 , @xmath76 is a long classical knot diagram , then @xmath74 is minimal . for a proof of theorem [ theorem2 ] we use following lemmas .
let @xmath77 be a number of long arc @xmath78 , where @xmath11 and @xmath12 are initial and final long arcs respectively
. then @xmath79s^{deg\,a}=$ ] @xmath80s^{deg\,a}+\sum_{a\subset { \gamma}_{+ } } \,[v_i : a]s^{deg\,a}$ ] .
consequently , @xmath81 , where @xmath82s^{deg\,a}$ ] , @xmath83 for @xmath84 .
thus , we have the natural decomposition of @xmath0-polynomial : @xmath85 , where @xmath86 . [ lem1 ] @xmath87 ; @xmath88 .
[ lem2 ] @xmath89/((p-1)(p - q),(q-1)(p - q))$ ] is zero divisor @xmath90 @xmath91 . by lemma [ lem1 ] @xmath92 @xmath93 @xmath94 , because @xmath95 .
consequently , @xmath96 if @xmath97 is not zero divisor .
it easy to check that @xmath98 , where @xmath99 denotes alexander polynomial .
it is known that @xmath100 .
hence , by lemma [ lem2 ] @xmath101 is not zero divisor , because @xmath102 . |
the discovery by donati et al .
( 2002 ) of a magnetic field in the periodic variable @xmath0orionisc ( hd37022 ; o67 v : morgan & keenan 1973 ; o7 : stahl et al .
1996 ) solidified the expectation that it is a hot analog of the magnetic b - type variables , the so called he - anomalous bp stars .
in contrast to the bp stars , the modulation of soft x - ray flux in @xmath0 ori c reveals the presence of a hot , co - rotating magnetosphere ( gagn et al . 1997 ) .
the magnetic and x - ray periods are presumed to correspond to the rotational period of 15.422 days , as determined from ultraviolet and optical line variations ( stahl et al .
1993 , stahl et al .
1996 , reiners et al .
nonetheless , questions have lingered as to how to reconcile the phased variations of strong optical and uv lines with the geometry of the bp magnetosphere model ( donati et al .
2002 ) .
the magnetic bp variables can be classified as a subgroup of a larger class of early - type stars that includes at least two pulsating @xmath3cephei stars ( henrichs et al . 1998 ) , the sn " stars4267 lines . ]
recently examined by shore et al .
( 2004 ) , and other related objects ( e.g. , smith & groote 2001 ; hereafter sg01 ) .
according to the usual picture , winds in magnetic bp stars are prevented from escaping from the magnetic poles , but are channeled outwards along dipolar field loops .
the wind is constrained to flow along these loops until it encounters streams from the opposite pole .
when the two streams collide , they shock , cool , and either circulate back to the star or leak out through the outer disk edge where they are lost to the system . the initial formulation of this picture by shore and collaborators ( e.g. , shore & brown 1990 ; hereafter sb90 ) envisaged polar outflows ( jets " ) and a cool , toroidal disk . subsequent magnetostatic modeling of the wind flows of bp stars by babel & montmerle ( 1997 ) has dealt with the post - shock cooling of the gas in the magnetospheric disk as the principal site of the uv and h@xmath4 line variations . in this formulation
, rotation is so small as to be negligible .
the frozen - in magnetosphere then resembles a wobbly inner tube that co - rotates with the bp star .
if the disk is extended enough in radius , it can contribute substantial line emission when it is viewed face - on .
moreover , disk regions passing in front of the star generally have a lower temperature than the photosphere and therefore remove additional flux from photospheric absorption lines . in the case of resonance lines ,
scattering of line flux has the same effect .
thus , over the cycle various viewing aspects of the wind / disk complex can produce a variety of emissions and absorptions in the uv resonance lines of , , and .
in addition , modulations of h@xmath4 emission ( short & bolton 1994 ) and high - level balmer line absorptions ( groote & hunger 1976 ) provide excellent diagnostics for well - developed magnetospheres , such as the one surrounding the prototypical he - strong star @xmath5orionise . along the hot edge of the classical bp domain , dense winds insure that the wind particles do not fractionate chemically , with the result that the surface abundances of these stars remain normal .
the study of the wind geometry of @xmath0 ori c is compelling because it is the only o star currently known to possess a magnetic field ( henrichs 2005 ) .
the presence of a strong field in a dense , high - velocity wind make its circumstellar environment a unique laboratory to study high - energy magnetohydrodynamic plasma ( babel & montmerle 1997 ) .
the interaction of various components of the wind produces the excitation of hot gas visible in x - ray spectra ( gagn et al .
2005 ) and potentially in the far - uv .
also , despite the inhibition caused by the strong magnetic field , the mass - loss rate of @xmath0 ori c is several hundred times larger than then mass - loss rates typical of hot bp stars ( donati et al . 2002 ) .
the strong stellar wind , together with the star s slow rotation rate , permit spectral features to be doppler shifted outside the photospheric line profile , which provides an opportunity to map the geometry of the wind over a wide range of stellar distances and azimuths .
in this study we adopt the values for the rotational and magnetic tilt angles @xmath6 and @xmath3 from the analysis of the longitudinal magnetic field curves by donati et al .
( 2002 ) : both parameters are close to 45 .
these values are the same as those adopted by stahl et al .
( 1996 ; hereafter s96 ) and nearly the same to those of reiners et al .
( 2000 ; hereafter r00 ) .
however , as noted by donati et al .
( 2002 ) , the mapping between line variations and magnetic phase adopted by s96 implies that the magnetic pole is viewed face - on at @xmath1 = 0.5 , which is difficult to reconcile with the observed h@xmath4 absorption at this phase . instead , the magnetic - field measurements of donati et al .
( 2002 ) show conclusively that phase @xmath1 = 0.0 corresponds to the passage of the north magnetic pole across the projected center of the star . at this phase , material in the equatorial plane
face - on " by a distant observer . at @xmath1 = 0.5
, the north magnetic pole has rotated to a point on the star s upper " limb and the south magnetic pole makes its only brief appearance at the star s lower " limb . at this time the observer sees the post - shock magnetosphere ( i.e. , the material near the magnetic equator ) edge - on . if the bp paradigm of sb90 and sg01 were applicable , maximum h@xmath4 emission and excess absorption in uv resonance lines should be observed at @xmath1 = 0 and 0.5 , respectively . as we will see below ,
this is not the case .
furthermore , donati et al . (
2002 ) were unable to explain the 50 kms@xmath7 blueshift in the centroid of the h@xmath4 emission component in the context of the bp picture .
the configuration described in this paper , shown as a sketch in figure[cartoon ] , is much different from this standard bp paradigm .
our geometry attributes the blueshifted absorption in the resonance lines at @xmath1 = 0.5 to a radially outflowing , high - velocity wind .
the point marked x " designates schematically the alfven radius at the magnetic equator .
for a very slowly rotating star this is where a bifurcation occurs in wind stream trajectories ( e.g. , ud - doula & owocki 2002 ) .
all streams sketched in fig.[cartoon ] are diverted toward the magnetic equator because the particles are constrained at first to flow along field lines by the dominant magnetic forces .
however , wind streams originating from high magnetic latitudes follow less confined loops that eventually extend beyond the alfven radius . as these loops traverse the alfven radius , radiation pressure ( which falls off less quickly with distance from the star ) overwhelms the magnetic forces and drags the field radially outward .
thus , these streams evolve to become a concentrated outflow confined to low magnetic latitudes .
in contrast , streams in the inner wind zone that originate from intermediate magnetic latitudes on the star remain confined and contribute or remove line flux only when they collide with one another , shock , and return to the star ( [ infall ] ) .
the fate of this wind component contrasts strongly with the formation of the quasi - static , corotating disk typical of bp stars , and is the subject of much of the rest of this paper .
in their seminal paper , s96 demonstrated that the emissions in h@xmath4 , @xmath84686 , and the absorptions of the and resonance doublets of @xmath0 ori c are regularly modulated over a period of 15.422 days .
this value is assumed to be the stellar rotational period .
phases in magnetic bp variables are traditionally reckoned from the time of magnetic maximum , i.e. , from the transit of a magnetic pole across the observer s line - of - sight . typically for these stars ,
if magnetic ephemerides are not available , the zero point is taken as the epoch of h@xmath4 emission maximum ( equivalent width minimum ) . because the central axis of the magnetosphere coincides with the magnetic plane , these zero points coincide .
using this zero point and period , s96 and r00 measured the equivalent - width variations with phase for several strong uv resonance and optical excited lines . the variations exhibit a maximum at @xmath1 = 0 , and both groups of authors interpreted this as excess absorptions in the line cores .
the variations in the uv resonance , h@xmath4 , h@xmath3 ( g. wade , private communication ) , and @xmath84686 lines are also highly nonsinusoidal .
as measured across the entire profile , the phased equivalent width variations have an m "- shape centered at @xmath1 = 0.5 .
the second maximum of the m " is centered at @xmath1 @xmath9 0.60.7 and is generally stronger than the first prong at @xmath1 = 0.4 . in the case of the h@xmath4 line ,
the local minimum at @xmath1 = 0.5 arises at least partly from the brief appearance of a red emission component ( stahl et al .
1993 ) . to reinvestigate the variations exhibited by the uv lines , particularly the and resonance doublets
, we retrieved the 22 high - dispersion _ iue _ spectra of @xmath0 ori c obtained through the large aperture with the swp camera from the multi - mission archive at space telescope .
the basic features of the variations are shown in figure [ grysc ] , which llustrates the fluctuations in as a function phase in in the form of a grayscale ( dynamic " ) spectrum .
the format of fig .
[ grysc ] is similar to figure 10 of s96 , except that the mean profile has been subtracted from individual spectra in the time series rather than the profile near @xmath10 , and the difference spectra illustrated in the image have been smoothed over 1 intervals .
[ grysc ] shows that the variations in the resonance doublet are dominated by enhanced absorption ( darker shades ) in the high - velocity region of the unusually shallow p cygni absorption trough .
however , weaker variations are also present throughout the absorption trough and even extend into the redshifted part of the profile .
the red - wing variations of the @xmath81548 line are confused by blending between the doublet components , but the corresponding variations of the @xmath81550 can be seen clearly in window
e " in the figure .
as discussed below , the weaker variations exhibit different behavior as a function of phase than the high - velocity absorption excess .
the equivalent width - phase curves ( e.g. , figure8 in s96 ) show some interesting features , and we have already mentioned the
m - shaped morphology .
these features , particularly those pertaining to the different dependences of fluxes across the line profile , can be investigated further by comparing the individual spectra at strategic phases .
a prime example of this is the different manner in which the fluxes across the line profiles contribute to the equivalent width of the full profile .
since the maximum swing in equivalent widths occurs over the comparatively narrow phase range between 0.6 and 0.9 , it is illuminating to compare the three available spectra at these phases .
these spectra are shown in figure [ c4mnmx ] .
the large flux variations at wavelengths shorter than 1543 are the most obvious differences .
however , we also note that the fluxes in the blue and red regions of the doublet vary in opposite senses .
notice that while profiles at @xmath1 = 0.50.6 exhibit increased blue absorption relative to the profiles observed near @xmath1 = 0.0 , the red - wing flux of the @xmath81550 line ( the region labeled `` e '' ) is high relative to the wing in the observation at @xmath1= 0.94 . put another way ,
the flux of the red wing of @xmath81550 is high when the fluxes of the ( blended ) blue wing of the doublet is low , and vice versa , over the rotational cycle .
the blue wing variations overwhelms the anticorrelated red - wing ones for the lines , which is why a depiction of the variation of the equivalent width extracted from the full profile conceals the red wing variation .
figure [ n5mnmx ] likewise shows observations of the @xmath111238 , 1242 at the same phases as the previous figure .
the blue - wing flux variations are again present , but in this case they are much smaller than for the doublet . in fact
, it is difficult to see a variation at all between phases 0.6 and 0.9 .
however , the red - wing variation , ( unblended in the two components ) , as indicated also by the smoothed difference spectrum at the bottom of the figure , is clearly visible .
another interesting feature of figures[c4mnmx ] and [ n5mnmx ] is that when allowance is made for the different local continuum levels , the amplitude of the red - wing variations is virtually the same for both components of the and doublets .
although there are additional complications due to blending , the implication is that the variable component is formed in an optically thick medium .
the phase - resolved fluctuations in different regions of the resonance doublet just discussed can be compared by plotting pseudo - equivalent widths extracted from different wavelength intervals .
to provide quantitative measures of the line absorptions , we computed equivalent width indices , " which are summed
net " fluxes in a defined series of windows that span the line .
we used the net fluxes to avoid the errors introduced by the ripple correction for high - dispersion _ iue _ data processed with newsips . " the equivalent width index in a given window was computed as the ratio of the total flux in it to the total flux in the particular order of the echelle spectrum .
like genuine equivalent widths , larger , positive values of these indices indicate greater absorption .
they can be measured with precision since they are independent of continuum placement errors , but they can not be converted to true equivalent widths in angstroms .
error bars were determined from the median point - to - point fluctuations in the values of the indices .
this criterion likely overestimates the uncertainties because of the slow variations of the index through the cycle .
when summed over the entire profile , our plots of the indices mimic the equivalent - width curves presented by s96
. however , the behavior in individual wavelength bins distributed across the weak p cygni profile is markedly different .
for example , the shortest wavelengths are dominated by a single , large - amplitude variation with maximum absorption near @xmath1 = 0.5 . in contrast , the indices formed from bins that sample the core and red wing reflect systematic , small variations in the red wing , as suggested by fig . [ c4mnmx ] .
these red - wing variations are illustrated in figure [ c4n5phas ] for and . from this figure
it is apparent that a sinusoidal variation ( shown as a dashed line for reference ) is not a good fit to the data .
[ c4n5phas ] also shows that the red - wing variations of the @xmath81242 profile ( 0+250 kms@xmath7 ) track the red - wing variations of the doublet .
we also find that a simple blue / red - profile dichotomy is an inadequate description of the variations of different regions of the doublet with phase . in figure[c4wng ] we represent fluxes in four additional 1-angstrom bins at increasingly blueshifted portions of the absorption trough . at the bottom of the figure we replot the curve from the red - wing extraction of 1550 in shown in fig.[c4n5phas ] . them "- pattern is present in curves _
b _ and _ c _ , but in curve _ a , _ representing the most negative velocities , the two prongs have merged together .
the same behavior is evident in the dynamic spectrum ( fig .
[ grysc ] ) . according to the revised geometry of the wind flow , spectroscopic observers in the right - hand section of fig .
[ cartoon ] will see the effects on the line profiles of a collimated , high - velocity wind rotating into their line of sight as phase @xmath1 = 0.5 approaches .
thus , the equivalent width indices extracted from the far - blue wing exhibit maximum absorption strength at this phase . the m "- shaped curves for the near - blue wing and smaller velocities can also be explained by this same flow geometry . between phases
@xmath1 = 0.30.4 and 0.60.7 , the absorptions occur at positions in the line that correspond approximately to the wind terminal velocity multiplied by the cosine of the angle between the line - of - sight to the center of the star and the bisector line in the equatorial plane .
since the column responsible for the high velocity absorption at @xmath1 = 0.5 is shifted to intermediate velocities just before and after this phase , a distant observer sees the maximum absorption at these times .
the wind flows seen against the rest of the stellar disk at these times are changing their direction and are also accelerating rapidly .
consequently , the absorptions these flows impose on the line profile are distributed over a broader velocity range and thus have a smaller influence on the equivalent widths extracted from any one narrow wavelength window in the absorption trough .
in table1 we summarize the relevant properties of a selection of optical and uv lines of @xmath12@xmath13oric reported in the recent literature .
all of these lines show a phase modulation .
we include the phases at which absorption equivalent widths of various lines attain their maximum absorption values .
those lines for which the m "- morphology is most prominent are denoted by phases 0.40.6 . "
we also show two entries for the @xmath111548,1550 doublet .
following s96 , the first of these refers to the equivalent width across the full profiles , while the second describes the results for the red half of the profile only .
a clear division can be made between the lines predominantly formed in the wind ( i.e. , the uv resonance lines , h@xmath4 , the emission in h@xmath3 , and @xmath84686 ) and the lines largely formed in the stellar photosphere ( e.g. , the excited transitions of and ) .
as already noted , h@xmath4 and @xmath84686 exhibit complicated emissions . at @xmath1 = 0.0
, a strong emission appears in the blue wing , and subtle variations interpreted by stahl et al .
( 1993 ) as emission appear at @xmath1 = 0.5 to the red of line center .
according to the sketch in figure[cartoon ] , blueshifted emissions might be observed as the magnetic pole crosses the projected center of the star at @xmath1 = 0 ; this possibility is addressed below .
for now we note that the appearance of strong h@xmath4 emission ( slightly blueshifted or otherwise , and discounting a nebular contribution ) is unusual in the spectra of mid - o dwarfs .
its presence in spectra of @xmath0 ori c is certainly attributable to the unusual conditions in the magnetically - channeled wind .
s96 and r00 illustrated the modulations of several excited lines of , , , and .
depending on the data quality , the absorption curves of these lines exhibit either a broad plateau centered at @xmath1 = 0.5 , or a shallow local
minimum at this phase .
although these authors constructed these curves by measuring the whole profile , differentiation with velocity can be found in the dynamic spectra of @xmath85592 and @xmath85812 presented in these papers .
these depictions show the blue - to - red migration of an apparent absorption , with line center crossing at @xmath14 .
the traditional interpretations for the origins of migrating subfeatures in line profiles of hot stars include surface spots due to chemical inhomogeneities , photospheric velocity fields due to nonradial pulsations , or circumstellar clouds forced into co - rotation .
the first of these alternatives is improbable in an o star having a chemically homogeneous wind .
the second is problematical for several reasons , but in particular because the migrating features are present during only half the cycle .
we will return presently to the third alternative , co - rotating clouds , " in the modified form of dense , infalling material .
s96 noted that the strength of the excited @xmath81371 line exhibits phase - dependent variations , but they did not discuss its behavior in detail .
figure [ o5phas ] shows the equivalent width indices from fluxes binned over the ranges from @xmath15300 to @xmath15100 kms@xmath7 and from @xmath15100 to @xmath16100 kms@xmath7 in the line profile .
these curves demonstrate that blue " and central - red " indices follow the morphology of the variations in the resonance lines ( fig .
[ c4n5phas ] ) ; i.e. , the two halves of the line profiles show anticorrelated behavior . thus , the blue half exhibits excess absorption at @xmath1 = 0.5 , probably because of the importance of the accelerating wind seen at this orientation . in contrast
, the red half shows either less absorption or more emission at this phase .
we have repeated these extractions for the excited @xmath81718 line .
although the curves are noisier , they show the same anticorrelated blue / red behavior .
we now return to the more complicated issue of the origin of the red - wing variations , e.g. , in the and resonance lines . within the limits imposed by the data quality ,
these and other optical lines listed in table1 show the same variations in the phase range 0.30.7 .
this fact suggests that the fluctuations are formed in the same geometrical structures in the wind . because the behavior of this ensemble of lines is similar regardless of their excitation and ionization potentials
, we expect that they are formed in a region of thermal equilibrium that is denser than an o - star wind .
the variations are probably optically thick , though further modeling is required to demonstrate this explicitly .
there are two basic interpretations for the red - wing variations : either they represent _ excess absorption _ at @xmath1 @xmath9 0.90.0 or they represent _ additional emission _ at @xmath1 @xmath9 0.50.6 .
the distribution of the circumstellar material is implicit in these interpretations .
in particular , if the red - wind variations are interpreted as excess emission , the material must be centered ( or nearly centered ) on either the star s magnetic pole or its magnetic equator in order to produce equivalent - width curves that are symmetric about the pole - on and equator - on phases .
we have identified three scenarios to address this issue : _ option i : _ : : central and red - wing _ absorption _ observed at @xmath1 = 0.0 in closed loops projected against the limb of the star .
the absorbing wind streams have a net positive velocity component relative to the observer . _
option ii : _ : : central and red - wing _ emission _ seen in the plane of the sky due to columns of material emanating from the magnetic poles . at @xmath1 = 0.5 these columns have rotated to opposite limbs of the star . _ option iii : _ : : central and red - wing _ emission _ from infalling material viewed edge - on in the magnetic equatorial plane .
_ option i _ suggests that the variations are due to absorption that is redshifted with respect to an observer situated above the magnetic pole in fig.[cartoon ] .
thus , the gas responsible for it must be moving with a vector component directed away from the observer , even though the wind is predominantly flowing outward along the magnetic field lines in this region .
this configuration is difficult to achieve , though it might arise from wind streams moving along closed loops near the star . in this case
, the line absorption would be visible at phases when the flow appears to cross the edge of the stellar disk .
to explore this idea , we can estimate the fraction of the stellar disk occulted by such a wind stream in the most ideal circumstances .
we have computed plasma absorptions using the hubeny atmosphere and spectral synthesis and cloud " transfer codes _ synspec _ and _ circus _ ( see sg01 and smith 2001 ) for a wind having a temperature of 35,00040,000k .
we find that in the
best " possible case for _ option i , _
( i.e. , for an optically thick wind with no background stellar - limb darkening ) , an absorption component formed at @xmath15200 kms@xmath7 would have to be produced by streams covering at least 9% of the stellar disk with a column density in h of at least 10@xmath17 @xmath18 in order to produce the redshifted variations observed in and .
even so , it is difficult to cover this much of the stellar disk with material that is redshifted by as much as @xmath19 kms@xmath7 ; see fig.[c4mnmx ] .
since the wind streams near the stellar limb are moving nearly parallel to the stellar surface , speeds of @xmath21500 kms@xmath7 are required to account for a projected component of 400 kms@xmath7 .
however , the terminal velocity of the wind in this star is no more than 2000 kms@xmath7 ( walborn & nichols 1994 ) : for a line - driven stellar wind characterized by a @xmath3=1 velocity law , a velocity as large as 1500 kms@xmath7 would only be reached at distances of @xmath24@xmath20 ; i.e. , far from the stellar surface . in view of the implausible degree of fine - tuning required by _
option i _ , we do not think it is likely .
_ option ii _ has the advantage that doppler - shifted emission is known to be present in some h and he lines at other phases , particularly @xmath1 = 0 , where it extends @xmath2300 kms@xmath7 redward of line center ( stahl et al .
1993 ; s96 ; g. wade , private communication ) .
since the behavior of the uv resonance lines is similar , it is reasonable to conclude that their variations are also due to changes in emission . according to _ option ii ,
_ scattering brings flux into our line of sight when the north magnetic pole rotates to the stellar limb .
an observer to the right in fig.[cartoon ] observes a glowing column of particles moving through the plane of the sky .
however , as the column ( jet " ) fans out from the plane of the sky , we expect to see primarily blueshifted h@xmath4 emission unless the line is optically thin .
thus , this picture fails to explain the exclusively redshifted emission observed in and .
perhaps the idea could be saved by invoking a significant departure from magnetic axi - symmetry , e.g. , with a decentered dipole and a jet that is compressed in one 180 azimuthal sector . in such a scheme
the observer might see more gas receding over a large area of the sky than gas exhibiting blueshifts in the foreground .
however , even this geometry would produce significant blueshifted emission , which is not indicated in figs.[grysc ] , [ c4mnmx ] or [ n5mnmx ] .
thus , _ option ii _ is probably not viable .
_ option iii _ associates the emission maximum at @xmath1 = 0.5 with infalling material near the magnetic equator rather than the poles .
since the lines of sight to this emission are also those that intersect the high - velocity wind moving toward the observer at this phase , this idea seems at first counterintuitive . how can the wind along the central line of sight to the star be approaching and receding at the same time ? furthermore
, this scenario must explain the appearance of redshifted in many strong lines , including h@xmath4 ( stahl et al .
1993 ) . in the resonance lines ,
the emission is probably optically thick .
moreover , since both low- and high - excitation lines show emission , it is most likely formed in a dense circumstellar medium and thus probably close to the star .
the geometry illustrated in fig.[cartoon ] addresses these issues , in particular by showing how rapidly moving , red- and blue - shifted material can be found between between the star and a distant observer at @xmath1 = 0.5 .
the presence of the infalling material required by _ option iii _ is is largely confirmed by recent magnetohydrodynamic ( mhd ) simulations by dr .
a. ud - doula ( private communication ; see gagn et al .
2005 for a description of this work ) , which also provide a physical basis for the geometry sketched in fig.[cartoon ] . as with previous studies of bp stars , the circulation results from the collision of wind streams that originate in opposite magnetic hemispheres and are guided by closed loops toward the magnetic equator . in the case of @xmath0 ori c , which has a much denser wind than a typical bp star , the collisions at the magnetic equator result in the formation of dense , optically thick blobs .
these structures can not be supported by radiation pressure from the star , and fall back toward the surface . since the blobs are still constrained to move along magnetic fields lines , they fall obliquely back toward the star along the lines that originally carried them to the magnetic equator .
observers viewing the system edge - on ( @xmath1 = 0.5 ) will see the infall as a redshifted , optically thick feature ; in spectra with sufficiently high signal - to - noise ratio , this emission might be variable from one cycle to the next . ud - doula s models also include effects of radiative cooling .
his models suggest that the condensations are part of the shocks .
these structures can be expected to contribute to line emission observed in the uv and optical regions .
since @xmath0 ori c rotates so slowly , it is appropriate to think of this circulation as a rigid , axisymmetric structure comprised of dense blobs .
we expect that the circulation we have described is related to the h@xmath4 flux and radial - velocity variations ( s96 ; gagn et al .
an additional consistency check for this infall picture is the _ absence _ of evidence for a cool , static disk in the magnetic equatorial plane .
such disks are the hallmarks of uv - variable bp stars .
our modeling of disks for early and late - type bp stars using _ synspec _ and _ circus _
( e.g. , sg01 ; smith 2003 ) suggests that the disk temperature is typically 0.6@xmath21 , which for @xmath0 ori c implies a disk temperature near 25,000k .
the ultraviolet line opacity tends to decrease with temperature in this temperature range . yet even for temperatures as high as 30,000k , our modeling experiments suggest that a geometrically thin disk with a column density of 10@xmath22 @xmath18 in h should be detectable as an absorption feature near rest velocity in the resonance doublet . the fact that a static disk is _ not _ visible implies that the circulating component of the wind interior to the alfven radius must be transported somewhere else .
we believe the redshifted emission at @xmath1=0.5 is evidence of this transport .
donati et al .
( 2002 ) noted that they were unable to duplicate the significantly blueshifted ( @xmath23 kms@xmath7 ) h@xmath4 emission in their simulation at the pole - on phase ( @xmath1 = 0.0 ) .
they found that the profile was formed mainly from a static disk , which in their model has a radius of @xmath23@xmath20 .
it is not surprising that the contribution from such a large disk would dominate the contribution from the polar wind , but the disk can not produce a blueshifted feature at @xmath1 = 0 in this geometry .
perhaps the blueshifts are caused by the acceleration of jet - like wind from the pole ?
we have investigated this possibility with _ circus .
_ we find that even with generous values for relevant parameters the contribution of a jet produces negligible h@xmath4 emission relative to the photospheric flux .
moreover , it is even more difficult for a jet - emission model to work for @xmath84686 .
in contrast , our same modeling experiments demonstrate that for similar temperatures and densities , a much smaller volume viewed beyond the projected face of the star could easily produce the observed h@xmath4 emission . from these results
we speculate that these blueshifts at @xmath1 = 0 have the same cause as the uv resonance line redshifts at @xmath1 = 0.5 : i.e. , they arise from the projection of the component of post - shock matter falling obliquely toward the star , as indicated by the closed arrowheads in fig.[cartoon ] .
from archival _ iue _ spectra of the magnetic o - type star @xmath0 ori c , we have discovered new variations of the and resonance doublets , as well as the strong , excited lines of @xmath81371 and @xmath81718 .
we found strong similarities between these variations and those exhibited by several prominent optical lines reported in the literature , including h@xmath4 and @xmath84686 .
lines with the largest variations exhibit an anticorrelation in blue- and red - wing fluxes at @xmath24 : as the red wing weakens , the blue wing strengthens .
in addition to this dichotomy , extractions of the profiles over different velocity bins show that the m "- shaped wave form present in optical lines is also present in the lines .
this morphology disappears at the highest wind velocities sampled ( i.e. , at the most blueshifted wavelengths ) , where a single - peaked absorption at @xmath1 = 0.5 replaces the m "- pattern .
the high - velocity strengthenings of the blue wing of at @xmath1 = 0.5 are explained by the flow geometry of ud - doula s mhd simulations .
these predict that radiative forces open the field lines and carry them away from the star once they extend beyond the alfven radius ( i.e. , several tenths of a stellar radius above the surface ; gagn et al . 2005 ) . at @xmath1 = 0.5 ,
the enhanced blue wings of these lines represent material accelerating away from the star along the magnetic plane .
we also argued that at @xmath25 and 0.7 the observer looks along a smaller projected path length through flows that have moderate outward velocities .
this means that the excess absorptions at these phases are weaker than the absorptions formed over the longer path length of the open , high - velocity flow viewed at @xmath1 = 0.5 ( fig.[cartoon ] ) .
the red variations at the edge - on phases ( @xmath26 ) , which we interpret as emission , require a more complicated geometry in which post - shock material flows towards the star , likely as dense clumps .
the presence of this material is also predicted by ud - doula s simulations , though in the current generation of models the infalling material only occurs episodically .
finally , _ in contrast to bona fide bp stars , there is no evidence for a cool , static disk . _
if post - shock material were to accumulate , given reasonable extrapolations from models of bp - star disks , a disk ought to be visible in spectra of @xmath0 ori c. we conclude that such accumulation does not happen in the circumstellar environment of @xmath0 ori c. the authors are grateful to drs .
marc gagn , richard townsend , asif ud - doula , and stan owocki for discussions of their mhd simulations which have led to a clearer description of our geometry .
we also thank both dr .
detlef groote and the referee for their careful readings of this paper and helpful comments to improve its clarity . finally , we thank dr .
gregg wade for permission to refer to his new optical data in advance of publication .
this work was supported in part by nasa grant # nng04ge75 g .
figure [ cartoon ] : : a sketch ( not to scale ) of the suggested wind / dipolar magnetic - field orientation for @xmath0 ori c at @xmath1 = 0.0 ( from top ) , 0.3 and 0.7 ( from 2 oclock position ) , and 0.5 ( from right ) , assuming angles @xmath6 = @xmath3 = 45 . in this planar projection
we have collapsed the observer s positions at @xmath1 = 0.3 and 0.7 , since they are at equal angles in and out of page , respectively .
the stellar wind expands freely only from the regions near the magnetic poles ( top and bottom of the star ) .
the magnetic field causes radiatively - driven material from other regions of the stellar surface to follow flux lines toward the magnetic equator .
the wind is divided into two zones according to its position relative to the star s alfven radius , which is designated by the dotted locus and the x " on the line - of - sight from the center of the star to the observer ( dashed line ) . beyond this radius
radiative forces dominate over magnetic forces and cause the wind to accelerate away from the star , approximately parallel to the magnetic equatorial plane . within the alfven radius , wind particles emerge from near the magnetic poles , follow closed loop lines ( open arrowheads ) , and encounter streams from the opposite hemisphere at the magnetic equator , where they shock and accumulate .
periodically , the particles formerly in the wind condense into optically thick blobs , which return to the star along the original field lines ( full arrowheads ) .
the redshifted variations suggest that the post - shock gas falls at speeds of 300400 kms@xmath7 .
figure [ grysc ] : : : dynamic spectrum of the 22 high - dispersion _ iue _ echellograms of @xmath0 ori c obtained through the large - aperture with the swp camera showing variations in the @xmath111548 , 1551 doublet as a function of magnetic / rotational phase .
the shading represents differences with respect to the time - averaged mean spectrum ( lower panel ) : darker shades denote phases and wavelengths where the profile is deeper ( i.e. , less flux ) than its time - averaged value .
the difference spectra have been smoothed over @xmath21 .
phase gaps in this spectrum are due to the paucity of observations .
brackets in the lower panel indicate the extent of the blue and red components of the doublet for an assumed terminal velocity of 2500 kms@xmath7 ( s96 ) .
although the dominant variations occur in the blue wing of the absorption trough ( window a " ) , significant variations occur throughout the profile .
the variations in window e " are anti - correlated with those in window
a. " figure [ c4mnmx ] : : : three high - resolution _ iue _ spectra in the region of the resonance lines for phases @xmath1 = 0.51 , 0.60 , and 0.94 ( swp54040 , swp07481 , and swp14665 ) . the units for the flux are ergs@xmath7@xmath18@xmath7 . all spectra are binned over 4 pixels , which undersamples the line cores and give them a sharp appearance .
the spectra obtained at @xmath1 = 0.60 and 0.94 ( solid and dashed lines , respectively ) exhibit the maximum difference in blue- and red - wing flux .
wavelength windows _ a e _ are referenced in figs.[grysc ] and [ c4wng ] .
following s96 , phase zero is defined by the epoch hjd=2448833.0 ( mjd=48832.5 ) .
figure [ n5mnmx ] : : : _ iue _ spectra for phases 0.60 and 0.64 ( swp07481 and 54058 ; solid line is their mean ) and 0.90 and 0.94 ( swp54094 and 14665 ; dashed line is their mean ) are compared in the region of the doublet of @xmath0 ori c. the spectra have been binned over 4 points .
the computed difference spectrum , binned over 8 points , is shown at the bottom in units of ergs@xmath7@xmath18@xmath7 .
figure [ c4n5phas ] : : : relative equivalent width index ( normalized to the maximum value ) versus phase for the red members of the and and doublets , as extracted over the velocity range between 0 and @xmath16250 kms@xmath7(01.3 ) . the dotted line is a reference sinusoid . figure [ c4wng ] : : : equivalent width index versus phase extracted from four velocity ranges in the blue wing of @xmath81548 ; see windows _ a e " _ in fig . [ grysc ] .
the curves have been offset vertically for clarity .
the corresponding extraction from 1548 was not used due to interference with the wind absorption of the 1550 line . note the progression from no variation , to an
m "- like pattern , and finally to a morphology featuring a single peak at @xmath1 @xmath9 0.5 .
figure [ o5phas ] : : : equivalent width index variations with phase of the excited @xmath81371 line , as extracted from negative and positive velocities of the line profile .
lccccl & 6563 & 10.2 & 13.2 & 0.40.6 & s96 - 4 + & 4686 & 48.4 & 54.4 & 0.40.6 & s96 - 5 + & 4471 & 21.0 & 24.6 & 0.0 & s96 - 13 + & 4541 & 51.0 & 54.4 & 0.0 & s96 - 14 + & 5801 & 37.5 & 64.5 & 0.0 & s96 - 15 + & 5812 & 37.5 & 64.5 & 0.0 & r00 - 7 + & 1548 , 1550 & 0.0 & 64.5 & 0.40.6 & s96 - 8 + & 1548 , 1550 & 0.0 & 64.5 & 0.0 & fig .
[ c4n5phas ] + & 1718 & 16.2 & 77.4 & 0.0 & [ uvexc ] + & 1238 , 1242 & 0.0 & 97.9 & 0.0 & fig .
[ c4n5phas ] + & 5592 & 33.9 & 55.0 & 0.0 & s96 - 16 + & 1371 & 19.7 & 113.9 & 0.0 & fig .
[ o5phas ] + & 1394 , 1402 & 0.0 & 45.1 & 0.40.6 & s96 - 9 + | the star @xmath0orionisc ( o67v ) is often cited as a hot analog of bp variables because its optical and uv line and x - ray continuum fluxes modulate over the magnetic / rotational period . in this circumstance ,
one expects emission and absorption components of the uv resonance lines to vary as a flattened magnetosphere co - rotates with the star . in this paper
we re - examine the detailed velocity behavior of several strong uv lines . whereas past work has focused on variations of the full profiles , we find that the blue and red wings of the and resonance lines exhibit anticorrelated modulations . these appear as absorption excesses at large blueshifts , and flux elevations at moderate redshifts at the edge - on phase @xmath1=0.5
. no rest - frame absorption features , which are the typical signatures of cool , static disks surrounding bp stars , can be detected at any phase .
we suggest that this behavior is caused by two geometrically distinct components of the wind , which are defined by the relationship between the extent of a magnetic loop and the local alfven radius .
streams on field lines opening outside this radius are first channeled toward the magnetic equator , but after reaching the alfven radius they are forced outward by radiative forces , eventually to become an expanding radial outflow .
this wind component causes blueshifted absorption as the co - rotating magnetic equatorial plane crosses the observer s line of sight ( @xmath1= 0.5 ) . the geometry of the inner component requires a more complicated interpretation .
wind streams first follow closed loops and collide at the magnetic equator with counterpart streams from the opposite pole .
there they coalesce and fall back to the star along their original field lines .
the high temperatures in these falling condensations cause the redshifted emission .
the rapid circulation of these flows is likely the reason for the absence of signatures of a cool disk ( e.g. , zero - velocity absorptions at @xmath1 @xmath2 0.5 ) in the strong uv lines . |
the one - dimensional spin-1/2 @xmath0 model is a quantum integrable system and has been well studied in statistical physics .
the hamiltonian is given by @xmath3 .
\label{eq : xxz_hamiltonian}\end{aligned}\ ] ] where @xmath4 with @xmath5 being pauli matrices . here
@xmath6 is the anisotropy parameter .
the eigenvalues and the eigenfunctions of the system were obtained by bethe ansatz @xcite , which allowed us to investigate many physical quantities @xcite . however , as for the correlation functions , it is not fully understood yet except for the @xmath7 case , where the system reduces to the free - fermion model @xcite . for general @xmath8 cases ,
nevertheless , there are also some recent progresses in the evaluation of the correlation functions , especially for the spin - spin correlation functions ( sscf ) @xmath9 and the emptiness formation probability ( efp ) @xmath10 @xcite . for example , via the field theoretical approach , we can now discuss the long - distance asymptotic behavior of the sscf @xcite , which decays algebraicly in the critical region @xmath11 .
the peculiar gaussian decay of the efp is also studied recently @xcite .
however these works contain some assumptions or approximations .
hence we usually need some other means , such as numerical simulations , to appeal their validity .
so it is more desirable if we can evaluate the correlations for an arbitrary @xmath12 exactly first , and derive its asymptotic behavior thereafter . at @xmath13 . ] in fact it has been known already that the analytical expressions for the sscf and the efp at arbitrary @xmath14 as well as other correlations among adjacent @xmath12-sites are given by certain @xmath12-dimensional integrals @xcite .
unfortunately there is a serious problem for these multiple integrals that it is hard to evaluate them , even numerically , because of the complicated structures of the integrand .
for this reason , it has been an important and challenging problem to simplify these multiple integral expressions systematically .
actually , a few years ago , boos _
et al _ succeeded in simplifying the multiple integrals for efp at @xmath15 .
they could express the efp up to @xmath16 as polynomials with respect to the specific values of @xmath17 function @xcite .
the same method was applied to the third - neighbor correlations at @xmath15@xcite .
furthermore , we have recently succeeded in generalizing the method to the general anisotropy parameter @xmath18 .
particularly we have reported in our previous papers , that the three - dimensional integrals relevant for the second neighbor correlations can be simplified to one - dimensional ones @xcite .
the main purpose of this paper is to present a detailed account of our method .
further , applying our method , we evaluate all the adjacent four point correlations including the third - neighbor correlation functions for general @xmath6 .
our significant finding is that these correlation functions are expressed as polynomials with respect to the following specific one - dimensional integrals , @xmath19 where the integration path is given by @xmath20 and the anisotropy parameter @xmath21 ( @xmath22 ) is either a real number from @xmath23 to @xmath24 or a purely imaginary number .
probably this property will be valid for any correlation functions for the @xmath25 chain ( cf .
we also remark the one - dimensional integrals ( [ ea : notation_first_intro ] ) and ( [ ea : notation_last_intro ] ) can be integrated analytically when @xmath21 takes a real rational number .
in such a case , we can obtain the complete analytic values for the correlation functions .
the outline of this paper is the following . in 2 , we show a general strategy to evaluate the multiple integral at the massless case . in 3 , we point out the similarity of the procedures in the massless case and the massive case . in 4 , we evaluate , as an example , an adjacent two - point correlation function by the strategy explained in 2 and 3 .
this derivation does not depend on @xmath18 , i.e. it works both in the massive and the massless cases . in 5 , we show the polynomial representation for the third - neighbor correlation functions and some of their analytical values at special points of @xmath6 .
the last section is devoted to the concluding remarks .
the outline of the derivation for an adjacent four - point correlation function is presented in appendix a. in appendix b , we list the polynomial expressions of all the independent correlation functions among up to four adjacent points .
in the massless region @xmath11 the correlation functions @xmath26 $ ] among adjacent @xmath27-sites are given by the following multiple integrals @xcite , @xmath28 \ : : = \ : \left < e^{\left(1\right)}_{\epsilon'_1 \epsilon_1 } \cdots e^{\left(n\right)}_{\epsilon'_n \epsilon_n } \right > \nonumber\\ & & = \nu^{-n\left(n-1\right)/2 } \prod_{a\in a_+ } \int_{c_+}\frac{-dx_a}{2\pi i } \prod_{a\in a_- } \int_{c_-}\frac{dx_a}{2\pi i } \prod_{1\leq l <
l'\leq n}\frac{\sinh\left(x_{l}-x_{l'}\right)}{\sinh\nu\left(x_{l}-x_{l'}-\pi i\right ) } \nonumber\\ & & { } \prod_{a\in a_+ } \frac { \sinh^{\bar a-1}\nu x_a \cdot \sinh^{n-\bar a}\nu \left(x_a-\pi i\right ) } { \sinh^n x_a } \prod_{a\in a_- } \frac { \sinh^{\bar a-1}\nu x_a \cdot \sinh^{n-\bar a}\nu \left(x_a+\pi i\right ) } { \sinh^n x_a}. \label{eq : integral_formula_p_n}\end{aligned}\ ] ] here @xmath29 and @xmath30 take either @xmath31 or @xmath32 , and @xmath33 are the @xmath34 matrix with 1 at the @xmath35-element and 0 elsewhere .
@xmath36 denote the operators acting as @xmath33 on the @xmath37-th spin site and as the identity on the other sites .
then the correlation function @xmath38 is defined by the expected value of the operator @xmath39 with respect to the ground state of the @xmath0 hamiltonian ( [ eq : xxz_hamiltonian ] ) .
for the description of the multiple integrals , we use the same notations as @xcite .
let us note them briefly below .
first the anisotropy parameter @xmath40 is defined by the relations @xmath41 .
as we are considering the massless region @xmath42 here , the parameter takes @xmath43 .
the contours @xmath44 in each integral goes from @xmath45 to @xmath46 .
the bars , e.g. @xmath47 , denotes the mapping from @xmath48 to @xmath48 defined by the following condition .
the @xmath31 s in the sequence @xmath49 are @xmath50 where @xmath51 is the number of @xmath32 in the sequence @xmath52 .
the sets @xmath53 and @xmath54 are defined as @xmath55 and @xmath56 respectively .
note that here we assume @xmath57 , because otherwise the expected value @xmath58 $ ] is equal to zero .
let us describe a strategy that may be used for the evaluation of general @xmath58 $ ] .
we first represent the integral formula ( [ eq : integral_formula_p_n ] ) as follows : @xmath59 \nonumber\\ & & = \nu^{-n\left(n-1\right)/2 } \prod_{j=1}^n \int_{c_-}\frac{dx_j}{2\pi i } u\left(x_1,\cdots , x_n\right ) t\left(\exp(2x_1 \nu),\cdots,\exp(2x_n \nu)\right ) , \label{eq : simplified_integral_form_p_n}\end{aligned}\ ] ] where @xmath60 @xmath61 and @xmath62 we can make a lot of simplification without evaluating integration but using only some simple observations and properties of the integrand of ( [ eq : simplified_integral_form_p_n ] ) . for this purpose , let us first define a `` weak '' equality @xmath63 .
we call that two functions @xmath64 and @xmath65 are `` weakly '' equivalent @xmath66 if @xmath67 & = & 0.\end{aligned}\ ] ] let us also introduce a `` canonical '' form of the function by the following formula @xmath68 } p_j^{(n)}\prod_{k=1}^j\frac1{y_{2k}-y_{2k-1 } } , \label{eq : define_t^c}\end{aligned}\ ] ] where @xmath69 are some polynomials with respect to @xmath70 such that @xmath71 and the coefficients @xmath72 are rational functions of @xmath73 .
our hypothesis is that for any @xmath74 and @xmath75 one can reduce the function @xmath76 defined by ( [ eq : define_t ] ) to the canonical form , i.e. there exist polynomials in ( [ eq : define_t^c ] ) such that @xmath77 unfortunately , for the moment we do not have a proof of this statement for any @xmath74 but we demonstrate how it works for @xmath78 in the following sections
. in fact , the problem of the evaluation of @xmath58 $ ] given by the integral ( [ eq : simplified_integral_form_p_n ] ) can be reduced to two steps .
the first step corresponds to the transformation into the `` canonical '' form .
the second step is the integration by means of this `` canonical '' form . for these procedures ,
we need several relations as follows . * * proposition 1 * + since the function @xmath79 is antisymmetric with respect to transposition of any pair of integration variables @xmath80 and @xmath81 , we have @xmath82 if the function @xmath83 is symmetric for at least one pair of @xmath84 .
therefore for an arbitrary function @xmath85 , we can transpose any pair of @xmath86 taking into consideration appearance of additional sign due to the antisymmetry of @xmath87 . for example , by transposing @xmath88 with @xmath70 , we get @xmath89 * * proposition 2 * + let the rational function @xmath90 have poles of the terms only @xmath91 and @xmath92 where @xmath93 is an integer , i.e. the product @xmath94 does not have poles of the terms @xmath95 .
then @xmath96 where @xmath14 is an integer and @xmath97 .
we can get four useful corollaries from the proposition 1 and 2 . *
_ corollary 1 _ @xmath98 where @xmath99 and the function @xmath100 does not depend on @xmath70 and @xmath101 .
the relation ( [ eq : reduce_pole ] ) also holds in the case @xmath88 is replaced with @xmath102 or @xmath103 .
+ _ proof _ @xmath104 at the two weak equalities above , we have used the proposition 1 . * _ corollary 2 _ @xmath105 where @xmath106 and the function @xmath107 does not depend on @xmath88 , @xmath70 , @xmath101 and @xmath108 . the relation ( [ eq : reduce_pole_2 ] ) is proved as follows .
@xmath109 here we used the proposition 1 similarly at the weak equality . *
_ corollary 3 _ @xmath110 where @xmath111 @xmath112 are integers and the function @xmath113 does not depend on @xmath88 and as above it is implied that @xmath97 . in our calculation ,
corollary 3 is used only in the case @xmath114 or @xmath14 .
+ _ proof _ + the relation ( [ eq : corollary_1 ] ) is derived easily from the relation @xmath115 @xmath116 , i.e. the proposition 2 . * _ corollary 4 _ @xmath117 where @xmath118 @xmath97 , @xmath119 and the function @xmath120 does not depend on @xmath70 and @xmath88 . in our calculation ,
corollary 4 is used only in the case @xmath121 or @xmath14 .
+ _ proof _
+ we get @xmath122 { } & & \frac { \left(y_j-1\right)^m y_j^p}{y_k - y_j } g\left(y_1,\cdots,\hat y_k , \cdots,\hat y_j,\cdots , y_n\right ) \nonumber\\ { } \makebox[-2 cm ] { } & \sim & -\frac { \left ( q y_j-1\right)^m \left ( q y_j\right)^p}{y_k- q y_j } g\left(y_1,\cdots,\hat y_k , \cdots,\hat y_j,\cdots ,
y_n\right ) \nonumber\\ { } \makebox[-2 cm ] { } & = & \left ( -\frac{\left ( y_k-1\right)^m y_k^p}{y_k- q y_j } + \frac{\left ( y_k-1\right)^m y_k^p -\left ( q y_j-1\right)^m \left ( q y_j\right)^p}{y_k- q y_j } \right ) g\left(y_1,\cdots,\hat y_k , \cdots,\hat y_j,\cdots ,
y_n\right ) \nonumber\\ { } \makebox[-2 cm ] { } & \sim & \left ( \frac{\left ( q y_k-1\right)^m \left ( q y_k\right)^p } { q\left(y_k - y_j\right ) } + \frac{\left ( y_k-1\right)^m y_k^p -\left ( q y_j-1\right)^m \left ( q y_j\right)^p}{y_k- q y_j } \right ) g\left(y_1,\cdots,\hat y_k , \cdots,\hat y_j,\cdots , y_n\right ) \nonumber\\ { } \makebox[-2 cm ] { } & \sim & \left ( \frac{\left ( q y_j-1\right)^m \left ( q y_j\right)^p } { q\left(y_k - y_j\right ) } + \frac{\left ( y_k-1\right)^m y_k^p -\left ( q y_j-1\right)^m \left ( q y_j\right)^p}{y_k- q y_j } \right ) g\left(y_1,\cdots,\hat y_k , \cdots,\hat y_j,\cdots , y_n\right ) . \nonumber\\ { } \makebox[-2cm]{}\end{aligned}\ ] ] the proposition 1 was used at the last weak equality , and the proposition 2 was used at the other weak equalities .
therefore , @xmath123 holds .
then , expanding both numerators according to the formulae : @xmath124 we arrive at the formula ( [ eq : corollary_2 ] ) .
the relations we wrote above are used to derive the `` canonical '' form of ( [ eq : general_integrant ] ) .
next , we show two elementary integral formulas .
* the first integral formula is @xmath125 where @xmath14 is an integer and @xmath126 . *
the second integral formula is @xmath127 where @xmath14 is an integer , @xmath128 is a real number and @xmath126 .
note that the l.h.s . of ( [ eq : value_of_integral_2 ] ) can be evaluated explicitly when @xmath14 is given explicitly .
for example , we have @xmath129 in the case @xmath130 and @xmath131 in the case @xmath132 , etc .. now let us consider the integrals of a special form , namely , the integral @xmath133 where @xmath134 is a polynomial of the form ( [ eq : define_p_j ] ) . using the relations
( [ eq : value_of_integral_1 ] ) and ( [ eq : value_of_integral_2 ] ) , we can integrate this expression @xmath135-times , namely with respect to @xmath136 and @xmath137 where @xmath138 , @xmath139 , while the variables @xmath140 are fixed . in this way
, we can evaluate the @xmath74-dimensional integral @xmath141 where @xmath142 is defined in ( [ eq : definition_u ] ) and @xmath143 is a canonical form function ( [ eq : define_t^c ] ) .
it results into a polynomial with respect to one - dimensional integrals where the coefficients are rational functions of @xmath144 and @xmath145 .
the correlation functions at massive case were obtained by jimbo _ et al _ @xcite . here
we use the representation in @xcite , @xmath146 : = \left < e^{\left(1\right)}_{\epsilon'_1 \epsilon_1 } \cdots e^{\left(n\right)}_{\epsilon'_n \epsilon_n } \right>\ : \nonumber \\ & = & \left(\phi i\right)^{-n(n-1)/2 } \prod_{j=1}^n \int_{c'}\frac{dx_j}{2\pi i } w\left(x_1,\cdots , x_n\right ) t\left(\exp(2x_1
\phi i),\cdots,\exp(2x_n \phi i)\right ) , \label{eq : simplified_integral_form_p_n_massive}\end{aligned}\ ] ] where @xmath147 is the integral path from @xmath148 to @xmath149 , @xmath150 is defined at ( [ eq : general_integrant ] ) , @xmath151 @xmath152 and @xmath153^{\frac { m(m+1)}2 - 1}.\end{aligned}\ ] ] the parameter @xmath154 is defined by the relations @xmath155 with @xmath156 . as in the massless case , we define weak equality @xmath63 .
let us say that two functions @xmath64 and @xmath65 are weakly equivalent : @xmath66 if @xmath157 & = & 0.\end{aligned}\ ] ] moreover , we use the term canonical " in the same way as in the massless case ( [ eq : define_t^c ] ) .
below we show the correlation function @xmath58 $ ] for the massive case can be evaluated in the same way as the massless case .
that is to say , we first obtain the canonical form , and then , reduce the multiple integral into the one - dimensional ones using this canonical form . in fact , all the equations in the massless case , i.e. ( [ eq : transposition])@xmath63([eq : weak_equation_last ] ) , also hold in the massive case .
the reason is that the function @xmath158 has the following properties : @xmath159 @xmath158 is an elliptic function with periods @xmath160 and @xmath161 with respect to each variable @xmath80 .
@xmath158 has zero points at @xmath162 and a pole only at @xmath163 , where @xmath164 .
the order of the pole is @xmath74 .
therefore , the canonical forms for the correlation function in the massive case are the same as those for the massless case .
next , we remark the relation @xmath165 holds .
the proof is as follows .
the both sides of ( [ eq : relation_w_u ] ) have following properties .
first , they are elliptic functions with periods @xmath161 and @xmath166 with respect to each variable @xmath80 .
second , the functions have poles only at @xmath167 and zeros at @xmath168 , where @xmath169 and @xmath170 are arbitrary integers and @xmath164 .
third , the order of the pole is @xmath74 and the order of any zero point is @xmath24 . finally , both values @xmath171 and @xmath172 are 1 . by the relation ( [ eq : relation_w_u ] )
, we find @xmath173 holds , where @xmath174 is the integral path from @xmath175 to @xmath176 and the both integrands have no pole on the integral paths respectively .
then , using the relation ( [ eq : value_of_integral_1 ] ) and ( [ eq : value_of_integral_2 ] ) , one can conclude , the multiple - integral @xmath177 where @xmath178 is defined as ( [ eq : definition_w ] ) and @xmath143 is a canonical form , reduces to a polynomial with respect to one - dimensional integral .
note that here the coefficients of the polynomial are rational functions of @xmath179 and @xmath180 .
we illustrate how the procedures in the previous sections work for the simplest case @xmath181 $ ] , i.e. , @xmath182 in the case @xmath183 , and @xmath184 in the case @xmath185 . here
@xmath186 and @xmath187 are defined in ( [ eq : definition_u ] ) and ( [ eq : definition_w ] ) respectively . for @xmath181 $ ] , it is very simple to perform the first step , namely , to get the `` canonical '' form ( [ eq : define_t^c ] ) because we do not need to reduce a power of denominator . indeed , @xmath188 where we have used the relations ( [ eq : transposition ] ) and ( [ eq : proposition1 ] ) .
then , using the relation ( [ eq : corollary_2 ] ) for @xmath130 , we have @xmath189 substituting it to the r.h.s . of ( [ eq : t_tmp_1 ] )
, we get @xmath190 and this is the `` canonical '' form for @xmath76 , i.e. the polynomials @xmath191 and @xmath192 at ( [ eq : define_t^c ] ) are equal to @xmath193 and @xmath194 respectively .
let us consider an integral : @xmath195 \nonumber\\ & = & \frac12 .
\label{eq : example_integral_1_massless}\end{aligned}\ ] ] this integrand corresponds to @xmath186 times the first term of ( [ eq : canonical_form_p2 ] ) .
here we used the formula ( [ eq : value_of_integral_1 ] ) at the second equality .
next , we evaluate another integral : @xmath196 % \nonumber\\ % & = & \frac1{2\pi^2\nu}\int_{-\infty}^{\infty}dy \frac{1}{\sinh y } \left ( \frac{e^{\nu y } } { \sinh\nu y } - \frac { i\pi\nu\left(q+1\right)\cosh\nu y}{\left(q-1\right ) \sinh\nu y } \right ) + \frac{\cosh y}{\sinh^2 y } \left ( \frac { i\pi\left(q+1\right)}{\left(q-1\right ) } - \frac { y e^{\nu y } } { \sinh\nu y } \right ) \nonumber\\ & = & % \int_{c_-}dy % \frac{1}{2\sinh y } % \left ( % - % \frac { i\left(q+1\right)\cosh\nu y}{\pi\left(q-1\right ) \sinh\nu y } % + % \frac { y } { \pi^2 \sinh^2\nu y } % \right ) % \nonumber\\ % & = & \int_{c_-}dy \frac{1}{\sinh y } \left ( - \frac{\cos \pi\nu}{2\pi\sin \pi\nu } \frac{\cosh\nu y } { \sinh\nu y } - \frac { 1 } { 2 \pi^2 } \frac\partial{\partial \nu}\frac{\cosh\nu y } { \sinh\nu y } \right ) .
\label{eq : example_integral_2_massless}\end{aligned}\ ] ] this integrand corresponds to @xmath186 times the second term of ( [ eq : canonical_form_p2 ] ) . at the second equality
, we used the formula ( [ eq : value_of_integral_2_1 ] ) , and at the third equality we used the relation @xmath197 where @xmath198 .
as for the massive case , by exchanging @xmath186 and @xmath199 with @xmath187 and @xmath147 , we can calculate as @xmath200 and @xmath201 actually , once we used the relation ( [ eq : relation_w_u_integral_form ] ) , we see the rest of the derivation is the same as the massless case . therefore , the final results of these two cases are very similar .
more precisely speaking , the results can be exchanged to each other with the exchange of @xmath202 and @xmath40 .
this property always holds in the evaluation of the other correlation functions . gathering all the results , we get the following expression @xmath203&=&\frac12 -\frac{\cos \pi\eta}{2\pi\sin \pi\eta } \int_{c_-}dy \frac{1}{\sinh y } \frac{\cosh\eta y } { \sinh\eta y } - \frac { 1 } { 2 \pi^2 } \int_{c_-}dy \frac{1}{\sinh y } \frac\partial{\partial \eta}\frac{\cosh\eta y } { \sinh\eta y},\end{aligned}\ ] ] where @xmath21 is equal to @xmath40 in the case @xmath204 and is equal to @xmath202 in the case @xmath205 .
using the method in the previous sections , we have analyzed the all correlation functions @xmath206 $ ] for @xmath207 .
actually the derivations of the canonical form from the multiple integral representations of the correlation functions are very similar to each other .
so , in [ sec : example ] , we explain the outline of its derivation for @xmath208 $ ] as an example .
also the rest of our tasks , i.e. the evaluation of the integrals with respect to the canonical forms , are straightforward if one uses the relations ( [ eq : value_of_integral_1 ] ) , ( [ eq : value_of_integral_2 ] ) and the integration by parts like ( [ eq : simple_relation ] ) . then , we omit the account of these calculations and give the final result in [ sec : results ] together with other independent results up to @xmath209 . the correlation functions which are not given in appendix b explicitly should be derived from the relations , @xmath210 \;=\ ; f[^ { \epsilon'_1 , \epsilon ' _ 2 , \cdots , \epsilon'_n,+ } _ { \epsilon _ 1 , \epsilon _ 2 , \cdots , \epsilon _ n,+ } ] + f[^ { \epsilon'_1 , \epsilon ' _ 2 , \cdots , \epsilon'_n,- } _ { \epsilon _ 1 , \epsilon _ 2 , \cdots , \epsilon _ n,- } ] , \\ &
f[^ { \epsilon'_1 , \epsilon ' _ 2 , \cdots , \epsilon'_n } _ { \epsilon _ 1 , \epsilon _ 2 , \cdots , \epsilon _
n } ] \;=\ ; f[^ { \epsilon'_n , \epsilon'_{n-1},\cdots , \epsilon'_1 } _ { \epsilon _ n , \epsilon _ { n-1},\cdots , \epsilon _ 1 } ] \;=\ ; f[^ { \epsilon _ 1 , \epsilon _ 2 , \cdots , \epsilon _
n } _ { \epsilon'_1 , \epsilon ' _ 2 , \cdots , \epsilon'_n } ] \;=\ ; f[^{-\epsilon'_1,-\epsilon ' _ 2 , \cdots,-\epsilon'_n } _ { -\epsilon _ 1,-\epsilon _ 2 , \cdots,-\epsilon _
n } ] , \\ & f[^ { \epsilon'_1 , \epsilon ' _ 2 , \cdots , \epsilon'_n } _ { \epsilon _ 1 , \epsilon _ 2 , \cdots , \epsilon _
n } ] \;=\;0\quad\quad\makebox{in case $ \#\{j|\epsilon'_j=+\}\neq\#\{j|\epsilon_j=+\}$}.\end{aligned}\ ] ] from our results , we can express many correlation functions as a polynomial with respect to one - dimensional integrals .
for example , the third - neighbor correlation functions @xmath211 and @xmath212 are given by @xmath213 here we used following notations to make the results short : @xmath214 where @xmath199 is an integral path from @xmath215 to @xmath216 and @xmath21 is equal to @xmath40 in case @xmath204 and is equal to @xmath202 in case @xmath217 .
note that we have used the relations
@xmath218 + f\left[^{++--}_{-+-+}\right ] , \\
\left < s_j^zs_{j+3}^z\right>&= & -2 f\left[^{+-+-}_{+-+-}\right ] + 2 f\left[^{++++}_{++++}\right ] -3 f\left[^{++}_{++}\right]+\frac34 , \end{aligned}\ ] ] to get the results ( [ eq : transverse_third_neighbor ] ) and ( [ eq : longitudinal_third_neighbor ] ) . the expressions ( [ eq : transverse_third_neighbor ] ) and ( [ eq : longitudinal_third_neighbor ] ) allow us to give the numerical values of the correlation functions with a very high precision .
they are plotted in fig .
[ graph : first ] .
furthermore , when @xmath21 is a real rational number , the one - dimensional integrals ( [ ea : notation_integral ] ) and ( [ ea : notation_last ] ) can be integrated analytically and we can get the complete analytical values for ( [ eq : transverse_third_neighbor ] ) and ( [ eq : longitudinal_third_neighbor ] ) . some explicit examples are given in table [ table : exact_values ] .
clll + @xmath40 & @xmath219 & @xmath212 & @xmath211 + + + @xmath220 & @xmath221 & @xmath222 & @xmath222 + & & @xmath223 & @xmath223 + + @xmath224 & @xmath23 & @xmath225 & @xmath226 + + @xmath227 & @xmath228 & @xmath229 & @xmath230 + + @xmath231 & @xmath232 & @xmath233 & @xmath234 + & & @xmath235 & @xmath236 + + @xmath237 & @xmath238 & @xmath239 & @xmath240 + & & & @xmath241 + + @xmath242 & @xmath243 & @xmath244 & @xmath245 + & @xmath246 & @xmath247 & @xmath248 + & & @xmath249 & @xmath250 + & & @xmath251 & @xmath252 + + @xmath253 & @xmath254 & @xmath255 & @xmath256 + + @xmath257 & @xmath258 & @xmath259 & @xmath260 + & & @xmath261 & @xmath262 + + @xmath24 & @xmath23 & @xmath23 & @xmath263 + + the polynomial representations of other correlation functions , @xmath264 etc .
, can be derived similarly .
here we simply show the shapes of the correlation functions in fig .
[ graph : second ] and fig .
[ graph : third ] by using the polynomial representations .
we remark that at @xmath15 our results reproduce the same values as in @xcite .
it is immediate to see them from the following asymptotic expansions at @xmath265 : @xmath266
we have analyzed multiple integral representations of correlation functions for @xmath0 model . in our results , we have given _ polynomial representations _
for all the adjacent points correlation functions under the restriction that the number of the points is one to four .
the _ polynomial _ means polynomial with respect to specific integrals ( [ ea : notation_first_intro ] ) and ( [ ea : notation_last_intro ] ) where the coefficients are rational functions of @xmath267 and @xmath268 . here
, @xmath268 is equal to @xmath18 , and @xmath21 is either a real number ( @xmath269 ) or a purely imaginary number ( @xmath270 ) .
it is intriguing that the correlation functions are given by the common expressions both in the massless regime and the massive regime .
probably it will be possible to generalize our results to more general @xmath271 model . on the other hand , we can , in principle , study the correlation functions even further for @xmath272 using our method . by surveying our results
, we conjecture that there exists _ polynomial representations _ for all these correlation functions , where only the one - dimensional integrals ( [ ea : notation_first_intro ] ) and ( [ ea : notation_last_intro ] ) with odd integer @xmath37 appear .
the authors are grateful to h. e. boos , k. hikami and m. wadati for valuable comments and stimulating discussions .
this work is in part supported by grant - in - aid for the scientific research ( b ) no .
14340099 and also by a 21st century coe program at tokyo technanometer - scale quantum physics " from the ministry of education , culture , sports , science and technology .
further gk is supported by the jsps research fellowships for young scientists and ms is supported by grant - in aid for young scientists ( b ) no .
14740228 . 9 h.a .
bethe , z. phys .
* 71 * ( 1931 ) 205 .
korepin , n.m .
bogoliubov and a.g .
izergin , _ quantum inverse scattering method and correlation functions _ , ( cambridge university press , cambridge , 1993 ) . m. takahashi , _ thermodynamics of one - dimensional solvable models _ , ( cambridge university press , cambridge , 1999 ) .
e. lieb , t. schultz and d. mattis , ann .
* 16 * ( 1961 ) 407 .
mccoy , phys .
* 173 * ( 1968 ) 531 .
m. shiroishi , m. takahashi and y. nishiyama , j. phys .
* 70 * ( 2001 ) 3535 .
n. kitanine , j.m .
maillet , n.a .
slavnov and v. terras , nucl .
b * 642 * ( 2002 ) 433 .
korepin , a.g .
izergin , f.h.l .
essler and d.b .
uglov , phys .
lett . a * 190 * ( 1994 ) 182 .
a. luther and i. peschel , phys .
b * 12 * ( 1975 ) 3908 .
s. lukyanov and a. zamolodchikov , nucl .
b * 493 * ( 1997 ) 571 .
s. lukyanov , phys .
b * 59 * ( 1999 ) 11163 .
s. lukyanov and v. terras , nucl .
b * 654 * ( 2003 ) 323 .
n. kitanine , j.m .
maillet , n.a .
slavnov and v. terras , j. phys .
a * 35 * ( 2002 ) l753 .
korepin , s. lukyanov , y. nishiyama and m. shiroishi , phys .
a * 312 * ( 2003 ) 21 .
n. kitanine , j.m .
maillet , n.a .
slavnov and v. terras , j. phys .
a * 35 * ( 2002 ) l385 .
n. kitanine , j.m .
maillet , n.a .
slavnov and v. terras , nucl .
b * 641 * ( 2002 ) 487 .
n. kitanine , j. m. maillet and v. terras , nucl .
b * 567 * ( 2000 ) 554 .
m. jimbo , k. miki , t. miwa and a. nakayashiki , phys .
lett . a * 168 * ( 1992 ) 256 .
m. jimbo and t. miwa , _ algebraic analysis of solvable lattice models _
( american mathematical society , providence , ri , 1995 ) a. nakayashiki , int .
j. mod . phys .
a * 9 * , ( 1994 ) 5673 .
m. jimbo and t. miwa , j. phys .
* 29 * ( 1996 ) 2923 .
boos and v.e .
korepin , j. phys . a : math .
gen * 34 * ( 2001 ) 5311 .
boos and v.e .
korepin , _ integrable models and beyond _ , edited by m. kashiwara and t. miwa ( birkhuser , boston , 2002 ) ; hep - th/0105144 .
boos , v.e .
korepin , y. nishiyama and m. shiroishi , j. phys . a : math .
gen * 35 * ( 2002 ) 4443 .
boos , v.e .
korepin and f.a .
smirnov , nucl .
b * 658 * ( 2003 ) 417 .
k. sakai , m. shiroishi , y. nishiyama and m. takahashi , phys .
e * 67 * ( 2003 ) 065101 .
g. kato , m. shiroishi , m. takahashi and k. sakai , j. phys .
* 36 * ( 2003 ) l337 .
m. takahashi , g. kato and m. shiroishi , j. phys .
soc . jpn .
* 73 * ( 2004 ) 245 .
boos , v.e .
korepin and f.a .
smirnov , hep - th/0305135 , to appear in moscow mathematical journal .
in this section , we derive canonical form of @xmath274 $ ] as an example . in the following ,
we assume @xmath275 as a polynomial with respect to @xmath276 , @xmath277 , @xmath278 and @xmath279 including negative power terms .
the polynomial @xmath76 with respect to @xmath274 $ ] are written as @xmath280 here the second equality is due to an elementary relation @xmath281 .
\nonumber\\ \label{eq : elementary_relation}\end{aligned}\ ] ] we named three terms in ( [ tmp : a_001 ] ) as @xmath282 , @xmath283 and @xmath284 order by order .
the two terms @xmath282 and @xmath284 are modified as @xmath285 at the two _ weak _ equalities above , we used the relation ( [ eq : reduce_pole ] ) .
we also used the elementary relation ( [ eq : elementary_relation ] ) at the two equalities above .
the term @xmath283 is modified as follows by use of elementary relations like ( [ eq : elementary_relation ] ) .
@xmath286 it is rather troublesome to show the further evaluation directly
. then , we introduce graphs which indicate certain sets of rational functions with respect to @xmath287 as follows .
each graph is made from three parts , i.e. circles , directed lines and non - directed lines .
first , there are four circles in a graph , which are placed at the four corners .
the circle at upper left ( right ) represents the variable @xmath276 ( @xmath277 ) , and the circle at lower left ( right ) represents the variable @xmath278 ( @xmath279 ) .
next , each graph contains directed lines or non - directed lines which connects two circles and indicates a certain analytic property of rational functions .
if a non - directed line connects @xmath88 and @xmath70 , it implies that the order of a pole at @xmath288 is @xmath24 at most . if @xmath88 and @xmath70 are connected by a directed line from @xmath88 to @xmath70 , the order of a pole at @xmath289 is @xmath24 at most , and at the same time the function has a factor @xmath290 with @xmath291 .
finally , we assume the rational functions do not have any other pole except for the one at @xmath292 . in this way
we can define a set of rational functions corresponding to each graph .
for example , @xmath293 is in the set @xmath294 we shall show the outline of further evaluation by means of these graphs . in fig .
[ fig : modification_stream_1 ] and [ fig : modification_stream_2 ] , some relations between the graphs are shown ; @xmath295 indicates that any function in `` graph1 '' can be modified into a sum of functions in `` graph2 '' using the week equality .
similarly , @xmath296 indicates that any function in `` graph1 '' can be modified into a sum of functions in `` graph2 '' and `` graph3 '' .
we can modify functions along the arrows between graphs in fig .
[ fig : modification_stream_1 ] and [ fig : modification_stream_2 ] .
below we prove it for some of the arrows .
first , let us consider the upper left arrow in fig .
[ fig : modification_stream_1 ] @xmath297 this modification can be proved primitively as follows .
@xmath298 where @xmath299 is a polynomial including negative powers . at the weak equality
, we used the relation ( [ eq : transposition ] ) .
next , we prove the relation @xmath300 which is one of arrows in fig . [
fig : modification_stream_1 ] .
@xmath301 here at the weak equality , we used the relation ( [ eq : transposition ] ) and ( [ eq : proposition1 ] ) .
the other arrows in fig . [ fig : modification_stream_1 ] are proved in the same way as the case ( [ eq : expample_arrow_2 ] ) . as for fig .
[ fig : modification_stream_2 ] , the left arrow is proved by the relation like ( [ eq : elementary_relation ] ) and ( [ eq : transposition ] ) , the middle arrow by ( [ eq : reduce_pole_2 ] ) and the right arrow by ( [ eq : reduce_pole ] ) . finally we note that any function in @xmath302 becomes canonical form using the relation ( [ eq : corollary_1 ] ) and ( [ eq : corollary_2 ] ) .
now , let us go back to modification of ( [ tmp : a_001 ] ) .
we note that any of functions @xmath303,@xmath304,@xmath305,@xmath306 is weekly equal to a function in the sets in the fig .
[ fig : modification_stream_1 ] .
then the discussion above indicates that @xmath307 can be modified into a canonical form according to the modification streams fig .
[ fig : modification_stream_1 ] and [ fig : modification_stream_2 ] with the relations ( [ eq : corollary_1 ] ) and ( [ eq : corollary_2 ] ) .
although we have actually come through the modifications , it is so tedious to show them here .
therefore , we show just the final result : @xmath308 where @xmath309
we use some notations defined in ( [ ea : notation_first])@xmath63([ea : notation_last ] ) . @xmath310 & = & \frac{1}{2 } \\ f\left [ \begin{array}{cc } + & + \\ + & + \end{array } \right ] & = & \frac{1}{2 } -\frac{{c_1 } } { 2 \pi { s_1}}{\zeta _ \eta}(1 ) -\frac{1}{2 { \pi } ^2}{\zeta'_\eta}(1 ) \\
f\left [ \begin{array}{cc } + & -\\
-&+ \end{array } \right ] & = & \frac{1}{2 \pi { s_1}}{\zeta _ \eta}(1 ) + \frac{{c_1}}{2 { \pi } ^2 } { \zeta'_\eta}(1 ) \\
f\left [ \begin{array}{ccc } + & + & + \\ + & + & + \end{array } \right ] & = & \frac{1}{2 } -\frac { 1 + 2 { c_2 } } { 2 \pi { s_2}}{\zeta _ \eta}(1 ) -\frac{3 } { 4 { \pi } ^2 } { \zeta'_\eta}(1 ) + \frac{3 { s_1 } } { 8 \pi { c_1}}{\zeta _ \eta}(3 ) + \frac { 1 - { c_2 } } { 16 { \pi } ^2 } { \zeta'_\eta}(3 )
\\ f\left [ \begin{array}{ccc } + & + & -\\
-&+&+ \end{array } \right ] & = & \frac{1}{2 \pi { s_2}}{\zeta _ \eta}(1 ) + \frac{{c_2 } } { 4 { \pi } ^2 } { \zeta'_\eta}(1 ) -\frac{3 { c_2 } \left ( 1 - { c_2 } \right ) } { 8 \pi { s_2 } } { \zeta _ \eta}(3 ) -\frac{1-{c_2 } } { 16 { \pi } ^2}{\zeta'_\eta}(3)\end{aligned}\ ] ] @xmath311 \nonumber\\ & = & \frac{1}{2 } -\frac{{c_2 } \left ( 5 + 6 { c_2 } \right ) } { 4 \pi { c_1 } { s_3}}{\zeta _ \eta}(1 ) -\frac{11 } { 12 { \pi } ^2}{\zeta'_\eta}(1 ) -\frac{21 + 61{c_2}+ 38 { c_4}+3 { c_6}+{c_8}}{32 \pi { c_1 } { s_3}}{\zeta _ \eta}(3 ) \nonumber\\ & & -\frac{5 + 6 { c_2}+3 { c_4}}{48 { \pi } ^2 } { \zeta'_\eta}(3 ) -\frac{5 \left ( 33 + 61 { c_2}+22 { c_4}+3 { c_6}+{c_8 } \right ) } { 64 \pi { c_1 } { s_3}}{\zeta _ \eta}(5 ) -\frac{7 + 2 { c_2}+{c_4}}{16 { \pi } ^2}{\zeta'_\eta}(5 ) \nonumber\\ & & + \frac{109 + 127 { c_2}+58 { c_4}+5 { c_6}+{c_8}}{32 { \pi } ^2 { s_1 } { s_3}}{\zeta _ \eta}(1 ) { \zeta _ \eta}(3 ) + \frac{{c_1 } \left ( 33 + 16 { c_2}+{c_4 } \right ) } { 16 { \pi } ^3 { s_1 } } { \zeta'_\eta}(1){\zeta _ \eta}(3 ) \nonumber\\ & & + \frac{{c_1 } \left ( 9+{c_4 } \right ) } { 16 { \pi } ^3 { s_1 } } { \zeta _ \eta}(1 ) { \zeta'_\eta}(3 ) + \frac{35 + 22 { c_2}+3 { c_4}}{96 { \pi } ^4}{\zeta'_\eta}(1 ) { \zeta'_\eta}(3 ) \nonumber\\ & & + \frac{5 \left ( 121 + 163 { c_2}+70 { c_4}+5 { c_6}+{c_8 } \right ) } { 64 { \pi } ^2 { s_1 } { s_3}}{\zeta _ \eta}(1 ) { \zeta _ \eta}(5 ) + \frac{5 { c_1 } \left ( 21 + 8 { c_2}+{c_4 } \right ) } { 16 { \pi } ^3 { s_1 } } { \zeta'_\eta}(1){\zeta _ \eta}(5 ) \nonumber\\ & & + \frac{{c_1 } \left ( 21 + 8 { c_2}+{c_4 } \right ) } { 16 { \pi } ^3 { s_1}}{\zeta _ \eta}(1 ) { \zeta'_\eta}(5 ) + \frac { 35 + 22 { c_2}+3 { c_4 } } { 32 { \pi } ^4}{\zeta'_\eta}(1 ) { \zeta'_\eta}(5 ) \nonumber\\ & & -\frac{3 \left ( 121 + 163 { c_2}+70 { c_4}+5 { c_6}+{c_8 } \right ) } { 128 { \pi } ^2 { s_1 } { s_3}}{{\zeta _ \eta}(3)}^2 -\frac{{c_1 } \left ( 21 + 8 { c_2}+{c_4 } \right ) } { 16 { \pi } ^3 { s_1}}{\zeta _ \eta}(3 ) { \zeta'_\eta}(3 ) \nonumber\\ & & -\frac{35 + 22 { c_2}+3 { c_4}}{192 { \pi } ^4}{{\zeta'_\eta}(3)}^2 \\ & & f\left [ \begin{array}{cccc } + & -&+&-\\ + & -&+&- \end{array } \right ] \nonumber\\ & = & \frac{{c_2 } } { 4 \pi { c_1 } { s_3 } } { \zeta _ \eta}(1 ) + \frac{1}{12 { \pi } ^2 } { \zeta'_\eta}(1 ) -\frac { 24 + 23 { c_2}+14 { c_4}+{c_6 } } { 16 \pi{c_1 } { s_3 } } { \zeta _ \eta}(3 ) \nonumber\\ & & -\frac { 4 + 3 { c_2 } } { 24 { \pi } ^2 } { \zeta'_\eta}(3 ) -\frac{5 \left ( 20 + 27 { c_2}+12 { c_4}+{c_6 } \right ) } { 32\pi { c_1 } { s_3 } } { \zeta _ \eta}(5 ) -\frac { 3 + 2 { c_2 } } { 8 { \pi } ^2 } { \zeta'_\eta}(5 ) \nonumber\\ & & + \frac{3 \left ( 4+{c_2 } \right ) \left ( 3 + 6 { c_2}+{c_4 } \right ) } { 16 { \pi } ^2 { s_1 } { s_3 } } { \zeta _ \eta}(1){\zeta _ \eta}(3 ) + \frac{5 { c_1 } \left ( 4+{c_2 } \right ) } { 8{\pi } ^3 { s_1 } } { \zeta'_\eta}(1){\zeta _ \eta}(3 ) \nonumber\\ & & + \frac{{c_1 } \left ( 4+{c_2 } \right ) } { 8{\pi } ^3 { s_1 } } { \zeta _ \eta}(1){\zeta'_\eta}(3 ) + \frac { 19 + 11 { c_2 } } { 48 { \pi } ^4 } { \zeta'_\eta}(1){\zeta'_\eta}(3 ) \nonumber\\ & & + \frac{15 \left ( 38 + 63 { c_2}+18 { c_4}+{c_6 } \right ) } { 64 { \pi } ^2 { s_1 } { s_3 } } { \zeta
_ \eta}(1){\zeta _ \eta}(5 ) + \frac{15 { c_1 } \left ( 4+{c_2 } \right)}{8{\pi } ^3 { s_1 } } { \zeta'_\eta}(1){\zeta _ \eta}(5 ) \nonumber\\ & & + \frac{3 { c_1 } \left ( 4+{c_2 } \right ) } { 8 { \pi } ^3 { s_1 } } { \zeta _ \eta}(1){\zeta'_\eta}(5 ) + \frac { 19 + 11 { c_2 } } { 16 { \pi } ^4 } { \zeta'_\eta}(1){\zeta'_\eta}(5 ) \nonumber\\ & & -\frac{9 \left ( 38 + 63 { c_2}+18 { c_4}+{c_6 } \right ) } { 128 { \pi } ^2 { s_1 } { s_3 } } { { \zeta _ \eta}(3)}^2 -\frac{3 { c_1 } \left ( 4+{c_2 } \right)}{8{\pi } ^3 { s_1 } } { \zeta _ \eta}(3){\zeta'_\eta}(3 ) \nonumber\\ & & -\frac { 19 + 11 { c_2 } } { 96 { \pi } ^4 } { { \zeta'_\eta}(3)}^2 \\ & & f\left [ \begin{array}{cccc } + & + & + & -\\ + & + & -&+ \end{array } \right ] \nonumber\\ & = & \frac{{c_2 } } { 2 \pi { s_3 } } { \zeta _ \eta}(1 ) + \frac{{c_1 } } { 6 { \pi } ^2 } { \zeta'_\eta}(1 ) + \frac { 11 + 41 { c_2}+9 { c_4}+{c_6 } } { 16 \pi { s_3 } } { \zeta _ \eta}(3 ) \nonumber\\ & & + \frac{{c_1 } \left ( 4 + 3 { c_2 } \right ) } { 24 { \pi } ^2 } { \zeta'_\eta}(3 ) + \frac{5 \left ( 17 + 33 { c_2}+9 { c_4}+{c_6 } \right)}{32 \pi { s_3 } } { \zeta _ \eta}(5 ) + \frac{{c_1 } \left ( 7 + 3 { c_2 } \right ) } { 16 { \pi } ^2 } { \zeta'_\eta}(5 ) \nonumber\\ & & -\frac{\left ( 4+{c_2 } \right ) \left ( 20 + 27 { c_2}+12 { c_4}+{c_6 } \right ) } { 16 { \pi } ^2 { s_2 } { s_3 } } { \zeta _ \eta}(1 ) { \zeta _ \eta}(3 ) -\frac { 49 + 48 { c_2}+3 { c_4 } } { 32 { \pi } ^3 { s_1 } } { \zeta'_\eta}(1){\zeta _
\eta}(3 ) \nonumber\\ & & -\frac { 11 + 8 { c_2}+{c_4 } } { 32 { \pi } ^3 { s_1 } } { \zeta _ \eta}(1 ) { \zeta'_\eta}(3 ) -\frac{{c_1 } \left ( 11 + 4 { c_2 } \right ) } { 24 { \pi } ^4 } { \zeta'_\eta}(1 ) { \zeta'_\eta}(3 ) \nonumber\\ & & -\frac{5 \left ( 219 + 328 { c_2}+148 { c_4}+24 { c_6}+{c_8 } \right ) } { 64 { \pi } ^2 { s_2 } { s_3 } } { \zeta _ \eta}(1 ) { \zeta _ \eta}(5 ) -\frac{5 \left ( 15 + 14 { c_2}+{c_4 } \right)}{16 { \pi } ^3 { s_1 } } { \zeta'_\eta}(1){\zeta _ \eta}(5 ) \nonumber\\ & & -\frac { 15 + 14 { c_2}+{c_4 } } { 16 { \pi } ^3 { s_1 } } { \zeta _ \eta}(1 ) { \zeta'_\eta}(5 ) -\frac{{c_1 } \left ( 11 + 4 { c_2 } \right ) } { 8 { \pi } ^4 } { \zeta'_\eta}(1 ) { \zeta'_\eta}(5 ) \nonumber\\ & & + \frac{3 \left ( 219 + 328 { c_2}+148 { c_4}+24 { c_6}+{c_8 } \right ) } { 128 { \pi } ^2 { s_2 } { s_3 } } { { \zeta _ \eta}(3)}^2 + \frac { 15 + 14 { c_2}+{c_4 } } { 16 { \pi } ^3 { s_1 } } { \zeta _ \eta}(3 ) { \zeta'_\eta}(3 ) \nonumber\\ & & + \frac{{c_1 } \left ( 11 + 4 { c_2 } \right ) } { 48 { \pi } ^4 } { { \zeta'_\eta}(3)}^2 \\ & & f\left [ \begin{array}{cccc } + & + & + & -\\ + & -&+&+ \end{array } \right ] \nonumber\\ & = & \frac{{c_1 } } { 2 \pi { s_3 } } { \zeta _ \eta}(1 ) + \frac{{c_2 } } { 6 { \pi } ^2 } { \zeta'_\eta}(1 ) + \frac{{c_1 } \left ( 19 + 3 { c_2}+9 { c_4 } \right ) } { 8 \pi { s_3 } } { \zeta _ \eta}(3 ) \nonumber\\ & & + \frac { 3 + 11 { c_2 } } { 48 { \pi } ^2 } { \zeta'_\eta}(3 ) + \frac{15 { c_1 } \left ( 2 + 2 { c_2}+{c_4 } \right ) } { 8 \pi { s_3 } } { \zeta _ \eta}(5 ) + \frac{{{c_1}}^2 \left ( 4+{c_2 } \right)}{8 { \pi } ^2 } { \zeta'_\eta}(5 ) \nonumber\\ & & -\frac{3 \left ( 28 + 51 { c_2}+18 { c_4}+3 { c_6 } \right ) } { 32 { \pi } ^2 { s_1 } { s_3 } } { \zeta _ \eta}(1 ) { \zeta _ \eta}(3 ) -\frac{{c_1 } \left ( 9+{c_2 } \right ) \left ( 3 + 2 { c_2 } \right)}{16 { \pi } ^3 { s_1 } } { \zeta'_\eta}(1 ) { \zeta _ \eta}(3 ) \nonumber\\ & & -\frac{5 { { c_1}}^3 } { 8 { \pi } ^3 { s_1 } } { \zeta _ \eta}(1 ) { \zeta'_\eta}(3 ) -\frac{5 { { c_1}}^2 \left ( 5+{c_2 } \right ) } { 48 { \pi } ^4 } { \zeta'_\eta}(1 ) { \zeta'_\eta}(3 ) \nonumber\\ & & -\frac{5 \left ( 106 + 177 { c_2}+66 { c_4}+11 { c_6 } \right ) } { 64 { \pi } ^2 { s_1 } { s_3 } } { \zeta _ \eta}(1 ) { \zeta _ \eta}(5 ) -\frac{5 { c_1 } \left ( 33 + 26 { c_2}+{c_4 } \right)}{32 { \pi } ^3 { s_1 } } { \zeta'_\eta}(1 ) { \zeta _ \eta}(5 ) \nonumber\\ & & -\frac{{c_1 } \left ( 33 + 26 { c_2}+{c_4 } \right)}{32 { \pi } ^3 { s_1 } } { \zeta _ \eta}(1 ) { \zeta'_\eta}(5 ) -\frac{5 { { c_1}}^2 \left ( 5+{c_2 } \right)}{16 { \pi } ^4 } { \zeta'_\eta}(1 ) { \zeta'_\eta}(5 ) \nonumber\\ & & + \frac{3 \left ( 106 + 177 { c_2}+66 { c_4}+11 { c_6 } \right ) } { 128 { \pi } ^2 { s_1 } { s_3 } } { { \zeta _ \eta}(3)}^2 + \frac{{c_1 } \left ( 33 + 26 { c_2}+{c_4 } \right)}{32 { \pi } ^3 { s_1 } } { \zeta _ \eta}(3 ) { \zeta'_\eta}(3 ) \nonumber\\ & & + \frac{5 { { c_1}}^2 \left ( 5+{c_2 } \right ) } { 96 { \pi } ^4 } { { \zeta'_\eta}(3)}^2 \\ & & f\left [ \begin{array}{cccc } + & + & -&+\\ + & -&+&+ \end{array } \right ] \nonumber\\ & = & \frac{1}{2 \pi { s_3 } } { \zeta _ \eta}(1 ) + \frac{{c_1 } } { 6 { \pi } ^2 } { \zeta'_\eta}(1 ) + \frac { 19 + 5 { c_2}+7 { c_4 } } { 8 \pi { s_3 } } { \zeta _ \eta}(3 ) \nonumber\\ & & + \frac{7 { c_1 } } { 24 { \pi } ^2 } { \zeta'_\eta}(3 ) + \frac{15 \left ( 2 + 2 { c_2}+{c_4 } \right ) } { 8 \pi { s_3 } } { \zeta _ \eta}(5 ) + \frac{{c_1 } \left ( 4+{c_2 } \right ) } { 8 { \pi } ^2 } { \zeta'_\eta}(5 ) \nonumber\\ & & -\frac{3 \left ( 28 + 51 { c_2}+18 { c_4}+3 { c_6 } \right)}{16 { \pi } ^2 { s_2 } { s_3 } } { \zeta _ \eta}(1 ) { \zeta _ \eta}(3 ) -\frac{\left ( 9+{c_2 } \right ) \left ( 3 + 2 { c_2 } \right ) } { 16 { \pi } ^3 { s_1 } } { \zeta'_\eta}(1 ) { \zeta _ \eta}(3 ) \nonumber\\ & & -\frac{5 { { c_1}}^2 } { 8 { \pi } ^3 { s_1 } } { \zeta _ \eta}(1 ) { \zeta'_\eta}(3 ) -\frac{5 { c_1 } \left ( 5+{c_2 } \right ) } { 48 { \pi } ^4 } { \zeta'_\eta}(1 ) { \zeta'_\eta}(3 ) \nonumber\\ & & -\frac{5 \left ( 106 + 177 { c_2}+66 { c_4}+11 { c_6 } \right)}{32 { \pi } ^2 { s_2 } { s_3 } } { \zeta _ \eta}(1 ) { \zeta _ \eta}(5 ) -\frac{5 \left ( 33 + 26 { c_2}+{c_4 }
\right ) } { 32 { \pi } ^3 { s_1 } } { \zeta'_\eta}(1 ) { \zeta _ \eta}(5 ) \nonumber\\ & & -\frac { 33 + 26 { c_2}+{c_4 } } { 32 { \pi } ^3 { s_1 } } { \zeta _ \eta}(1 ) { \zeta'_\eta}(5 ) -\frac{5 { c_1 } \left ( 5+{c_2 } \right ) } { 16 { \pi } ^4 } { \zeta'_\eta}(1 ) { \zeta'_\eta}(5 ) \nonumber\\ & & + \frac{3 \left ( 106 + 177 { c_2}+66 { c_4}+11 { c_6 } \right ) } { 64 { \pi } ^2 { s_2 } { s_3 } } { { \zeta _ \eta}(3)}^2 + \frac { 33 + 26 { c_2}+{c_4 } } { 32 { \pi } ^3 { s_1 } } { \zeta _ \eta}(3 ) { \zeta'_\eta}(3 ) \nonumber\\ & & + \frac{5 { c_1 } \left ( 5+{c_2 } \right)}{96 { \pi } ^4 } { { \zeta'_\eta}(3)}^2 \\ & & f\left [ \begin{array}{cccc } + & + & + & -\\
-&+&+&+ \end{array } \right ] \nonumber\\ & = & \frac{1}{2 \pi { s_3 } } { \zeta _ \eta}(1 ) -\frac{{c_1 } \left ( 1 - 2 { c_2 } \right ) } { 6 { \pi } ^2 } { \zeta'_\eta}(1 ) + \frac { 15 + 7 { c_2}+5 { c_4}+4 { c_6 } } { 8 \pi { s_3 } } { \zeta _ \eta}(3 ) \nonumber\\ & & -\frac{{c_1 } \left ( 1 - 8 { c_2 } \right ) } { 24 { \pi } ^2 } { \zeta'_\eta}(3 ) + \frac{5 \left ( 5 + 6 { c_2}+3 { c_4}+{c_6 } \right ) } { 8 \pi { s_3 } } { \zeta _ \eta}(5 ) + \frac{{c_1 } \left ( 5 + 4 { c_2}+{c_4 } \right ) } { 16 { \pi } ^2 } { \zeta'_\eta}(5 ) \nonumber\\ & & -\frac{{c_1 } \left ( 14 + 48 { c_2}+9 { c_4}+4 { c_6 } \right ) } { 8 { \pi } ^2 { s_1 } { s_3 } } { \zeta _ \eta}(1 ) { \zeta _ \eta}(3 ) -\frac { 34 + 53 { c_2}+12 { c_4}+{c_6 } } { 32 { \pi } ^3 { s_1 } } { \zeta'_\eta}(1 ) { \zeta _ \eta}(3 ) \nonumber\\ & & -\frac { 5 + 3 { c_2}+2 { c_4 } } { 16 { \pi } ^3 { s_1 } } { \zeta _ \eta}(1 ) { \zeta'_\eta}(3 ) -\frac{{{c_1}}^3 \left ( 11 + 4 { c_2 } \right)}{24 { \pi } ^4 } { \zeta'_\eta}(1 ) { \zeta'_\eta}(3 ) \nonumber\\ & & -\frac{5 { c_1 } \left ( 23 + 48 { c_2}+15 { c_4}+4 { c_6 } \right ) } { 16 { \pi } ^2 { s_1 } { s_3 } } { \zeta _ \eta}(1 ) { \zeta _ \eta}(5 ) -\frac{5 { { c_1}}^2 \left ( 15 + 14 { c_2}+{c_4 } \right)}{16 { \pi } ^3 { s_1 } } { \zeta'_\eta}(1 ) { \zeta _ \eta}(5 ) \nonumber\\ & & -\frac{{{c_1}}^2 \left ( 15 + 14 { c_2}+{c_4 } \right ) } { 16 { \pi } ^3 { s_1 } } { \zeta _ \eta}(1 ) { \zeta'_\eta}(5 ) -\frac{{{c_1}}^3 \left ( 11 + 4 { c_2 } \right ) } { 8 { \pi } ^4 } { \zeta'_\eta}(1 ) { \zeta'_\eta}(5 ) \nonumber\\ & & + \frac{3 { c_1 } \left ( 23 + 48 { c_2}+15 { c_4}+4 { c_6 } \right ) } { 32 { \pi } ^2 { s_1 } { s_3 } } { { \zeta _ \eta}(3)}^2 + \frac{{{c_1}}^2 \left ( 15 + 14 { c_2}+{c_4 } \right ) } { 16 { \pi } ^3 { s_1 } } { \zeta _ \eta}(3 ) { \zeta'_\eta}(3 ) \nonumber\\ & & + \frac{{{c_1}}^3 \left ( 11 + 4 { c_2 } \right ) } { 48 { \pi } ^4 } { { \zeta'_\eta}(3)}^2 \\ & & f\left [ \begin{array}{cccc } + & + & -&-\\
-&+&-&+ \end{array } \right ] \nonumber\\ & = & \frac{1}{4 \pi { s_3 } } { \zeta _ \eta}(1 ) -\frac{{c_1 } \left ( 1 - 2 { c_2 } \right ) } { 12 { \pi } ^2 } { \zeta'_\eta}(1 ) -\frac{{c_2 } \left ( 19 + 12 { c_2 } \right ) } { 8 \pi { s_3 } } { \zeta _ \eta}(3 ) \nonumber\\ & & -\frac{{c_1 } \left ( 5 + 2 { c_2 }
\right)}{24 { \pi } ^2 } { \zeta'_\eta}(3 ) -\frac{5 \left ( 8 + 15 { c_2}+6 { c_4}+{c_6 } \right ) } { 16 \pi { s_3 } } { \zeta _ \eta}(5 ) -\frac{{c_1 } \left ( 2 + 3 { c_2 } \right ) } { 8 { \pi } ^2 } { \zeta'_\eta}(5 ) \nonumber\\ & & + \frac{{c_1 } \left ( 56 + 57 { c_2}+36 { c_4}+{c_6 } \right ) } { 16 { \pi } ^2 { s_1 } { s_3 } } { \zeta _ \eta}(1 ) { \zeta _ \eta}(3 ) + \frac { 43 + 46 { c_2}+11 { c_4 } } { 32 { \pi } ^3 { s_1 } } { \zeta'_\eta}(1 ) { \zeta _ \eta}(3 ) \nonumber\\ & & + \frac { 5 + 14 { c_2}+{c_4 } } { 32 { \pi } ^3 { s_1 } } { \zeta _ \eta}(1 ) { \zeta'_\eta}(3 ) + \frac{{{c_1}}^3 \left ( 14+{c_2 } \right)}{24 { \pi } ^4 } { \zeta'_\eta}(1 ) { \zeta'_\eta}(3 ) \nonumber\\ & & + \frac{15 { c_1 } \left ( 20 + 27 { c_2}+12 { c_4}+{c_6 } \right ) } { 32 { \pi } ^2 { s_1 } { s_3 } } { \zeta _ \eta}(1 ) { \zeta _ \eta}(5 ) + \frac{15 { { c_1}}^2 \left ( 3 + 2 { c_2 } \right ) } { 8 { \pi } ^3 { s_1 } } { \zeta'_\eta}(1 ) { \zeta _
\eta}(5 ) \nonumber\\ & & + \frac{3 { { c_1}}^2 \left ( 3 + 2 { c_2 } \right)}{8 { \pi } ^3 { s_1 } } { \zeta _ \eta}(1 ) { \zeta'_\eta}(5 ) + \frac{{{c_1}}^3 \left ( 14+{c_2 } \right)}{8 { \pi } ^4 } { \zeta'_\eta}(1 ) { \zeta'_\eta}(5 ) \nonumber\\ & & -\frac{9 { c_1 } \left ( 20 + 27 { c_2}+12 { c_4}+{c_6 } \right ) } { 64 { \pi } ^2 { s_1 } { s_3 } } { { \zeta _ \eta}(3)}^2 -\frac{3 { { c_1}}^2 \left ( 3 + 2 { c_2 } \right ) } { 8 { \pi } ^3 { s_1 } } { \zeta _ \eta}(3 ) { \zeta'_\eta}(3 ) \nonumber\\ & & -\frac{{{c_1}}^3 \left ( 14+{c_2 } \right)}{48 { \pi } ^4 } { { \zeta'_\eta}(3)}^2 \\ & & f\left [ \begin{array}{cccc } + & + & -&-\\
-&-&+&+ \end{array } \right ] \nonumber\\ & = & \frac{1}{4 \pi { c_1 } { s_3 } } { \zeta _ \eta}(1 ) + \frac{{c_4}}{12 { \pi } ^2 } { \zeta'_\eta}(1 ) -\frac { 6 + 15 { c_2}+10 { c_4 } } { 8 \pi { c_1 } { s_3 } } { \zeta _ \eta}(3 ) \nonumber\\ & & -\frac { 6+{c_4 } } { 24 { \pi } ^2 } { \zeta'_\eta}(3 ) -\frac{5 { c_2 } \left ( 11 + 15 { c_2}+3 { c_4}+{c_6 } \right ) } { 16 \pi { c_1 } { s_3 } } { \zeta _ \eta}(5 ) -\frac { 2 + 2 { c_2}+{c_4 } } { 8 { \pi } ^2 } { \zeta'_\eta}(5 ) \nonumber\\ & & + \frac { 46 + 57 { c_2}+36 { c_4}+11 { c_6 } } { 16 { \pi } ^2 { s_1 } { s_3 } } { \zeta _ \eta}(1 ) { \zeta _ \eta}(3 ) + \frac{{c_1 } \left ( 12 + 9 { c_2}+4 { c_4 } \right ) } { 8 { \pi } ^3 { s_1 } } { \zeta'_\eta}(1 ) { \zeta _ \eta}(3 ) \nonumber\\ & & + \frac{5 { c_1 } { c_2 } } { 8 { \pi } ^3 { s_1 } } { \zeta _ \eta}(1 ) { \zeta'_\eta}(3 ) + \frac{{{c_1}}^2 \left ( 13 + 16 { c_2}+{c_4 } \right)}{48 { \pi } ^4 } { \zeta'_\eta}(1 ) { \zeta'_\eta}(3 ) \nonumber\\ & & + \frac{5 \left ( 51 + 76 { c_2}+40 { c_4}+12 { c_6}+{c_8 } \right ) } { 32 { \pi } ^2 { s_1 } { s_3 } } { \zeta _ \eta}(1 ) { \zeta _ \eta}(5 ) + \frac{5 { c_1 } \left ( 6 + 7 { c_2}+2 { c_4 } \right ) } { 8 { \pi } ^3 { s_1 } } { \zeta'_\eta}(1 ) { \zeta _ \eta}(5 ) \nonumber\\ & & + \frac{{c_1 } \left ( 6 + 7 { c_2}+2 { c_4 } \right)}{8 { \pi } ^3 { s_1 } } { \zeta _ \eta}(1 ) { \zeta'_\eta}(5 ) + \frac{{{c_1}}^2 \left ( 13 + 16 { c_2}+{c_4 } \right)}{16 { \pi } ^4 } { \zeta'_\eta}(1 ) { \zeta'_\eta}(5 ) \nonumber\\ & & -\frac{3 \left ( 51 + 76 { c_2}+40 { c_4}+12 { c_6}+{c_8 } \right ) } { 64 { \pi } ^2 { s_1 } { s_3 } } { { \zeta _ \eta}(3)}^2 -\frac{{c_1 } \left ( 6 + 7 { c_2}+2 { c_4 } \right ) } { 8 { \pi } ^3 { s_1 } } { \zeta _ \eta}(3 ) { \zeta'_\eta}(3 ) \nonumber\\ & & -\frac{{{c_1}}^2 \left ( 13 + 16 { c_2}+{c_4 } \right)}{96 { \pi } ^4 } { { \zeta'_\eta}(3)}^2 \\ & & f\left [ \begin{array}{cccc } + & -&+&-\\ -&+&-&+ \end{array } \right ] \nonumber\\ & = & \frac{1}{4 \pi { c_1 } { s_3 } } { \zeta _ \eta}(1 ) + \frac{{c_2}}{12 { \pi } ^2 } { \zeta'_\eta}(1 ) -\frac { 11 + 37 { c_2}+13 { c_4}+{c_6 } } { 16 \pi { c_1 } { s_3 } } { \zeta _ \eta}(3 ) \nonumber\\ & & -\frac { 3 + 4 { c_2 } } { 24 { \pi } ^2 } { \zeta'_\eta}(3 ) -\frac{5 \left ( 8 + 15 { c_2}+6 { c_4}+{c_6 } \right)}{16 \pi { c_1 } { s_3 } } { \zeta _ \eta}(5 ) -\frac { 2 + 3 { c_2 } } { 8 { \pi } ^2 } { \zeta'_\eta}(5 ) \nonumber\\ & & + \frac { 55 + 60 { c_2}+33 { c_4}+2 { c_6 } } { 16 { \pi } ^2 { s_1 } { s_3 } } { \zeta _ \eta}(1 ) { \zeta _ \eta}(3 ) + \frac{5 { c_1 } \left ( 3 + 2 { c_2 } \right ) } { 8 { \pi } ^3 { s_1 } } { \zeta'_\eta}(1 ) { \zeta _
\eta}(3 ) \nonumber\\ & & + \frac{{c_1 } \left ( 3 + 2 { c_2 } \right ) } { 8 { \pi } ^3 { s_1 } } { \zeta _ \eta}(1 ) { \zeta'_\eta}(3 ) + \frac{{{c_1}}^2 \left ( 14+{c_2 } \right ) } { 24 { \pi } ^4 } { \zeta'_\eta}(1 ) { \zeta'_\eta}(3 ) \nonumber\\ & & + \frac{15 \left ( 20 + 27 { c_2}+12 { c_4}+{c_6 } \right ) } { 32 { \pi } ^2 { s_1 } { s_3 } } { \zeta _ \eta}(1 ) { \zeta _ \eta}(5 ) + \frac{15 { c_1 } \left ( 3 + 2 { c_2 } \right ) } { 8 { \pi } ^3 { s_1 } } { \zeta'_\eta}(1 ) { \zeta _ \eta}(5 ) \nonumber\\ & & + \frac{3 { c_1 } \left ( 3 + 2 { c_2 } \right ) } { 8 { \pi } ^3 { s_1 } } { \zeta _ \eta}(1 ) { \zeta'_\eta}(5 ) + \frac{{{c_1}}^2 \left ( 14+{c_2 } \right)}{8 { \pi } ^4 } { \zeta'_\eta}(1 ) { \zeta'_\eta}(5 ) \nonumber\\ & & -\frac{9 \left ( 20 + 27 { c_2}+12 { c_4}+{c_6 } \right ) } { 64 { \pi } ^2 { s_1 } { s_3 } } { { \zeta _ \eta}(3)}^2 -\frac{3 { c_1 } \left ( 3 + 2 { c_2 } \right)}{8 { \pi } ^3 { s_1 } } { \zeta _ \eta}(3 ) { \zeta'_\eta}(3 ) \nonumber\\ & & -\frac{{{c_1}}^2 \left ( 14+{c_2 } \right ) } { 48 { \pi } ^4 } { { \zeta'_\eta}(3)}^2 \\ & & f\left [ \begin{array}{cccc } + & -&-&+\\ -&+&+&- \end{array } \right ] \nonumber\\ & = & \frac{1}{4 \pi { c_1 } { s_3 } } { \zeta _ \eta}(1 ) + \frac{1}{12 { \pi } ^2 } { \zeta'_\eta}(1 ) -\frac { 5 + 18 { c_2}+7 { c_4}+{c_6 } } { 8 \pi { c_1 } { s_3 } } { \zeta _ \eta}(3 ) \nonumber\\ & & -\frac { 1 + 6 { c_2 } } { 24 { \pi } ^2 } { \zeta'_\eta}(3 ) -\frac{5 \left ( 8 + 15 { c_2}+6 { c_4}+{c_6 } \right)}{16 \pi { c_1 } { s_3 } } { \zeta _ \eta}(5 ) -\frac { 2 + 3 { c_2 } } { 8 { \pi } ^2 } { \zeta'_\eta}(5 ) \nonumber\\ & & + \frac { 55 + 60 { c_2}+33 { c_4}+2 { c_6 } } { 16 { \pi } ^2 { s_1 } { s_3 } } { \zeta _ \eta}(1 ) { \zeta _ \eta}(3 ) + \frac{5 { c_1 } \left ( 3 + 2 { c_2 } \right ) } { 8 { \pi } ^3 { s_1 } } { \zeta'_\eta}(1 ) { \zeta _ \eta}(3 ) \nonumber\\ & & + \frac{{c_1 } \left ( 3 + 2 { c_2 } \right)}{8 { \pi } ^3 { s_1 } } { \zeta _ \eta}(1 ) { \zeta'_\eta}(3 ) + \frac{{{c_1}}^2 \left ( 14+{c_2 } \right)}{24 { \pi } ^4 } { \zeta'_\eta}(1 ) { \zeta'_\eta}(3 ) \nonumber\\ & & + \frac{15 \left ( 20 + 27 { c_2}+12 { c_4}+{c_6 } \right ) } { 32 { \pi } ^2 { s_1 } { s_3 } } { \zeta _ \eta}(1 ) { \zeta _ \eta}(5 ) + \frac{15 { c_1 } \left ( 3 + 2 { c_2 } \right ) } { 8 { \pi } ^3 { s_1 } } { \zeta'_\eta}(1 ) { \zeta _ \eta}(5 ) \nonumber\\ & & + \frac{3 { c_1 } \left ( 3 + 2 { c_2 } \right)}{8 { \pi } ^3 { s_1 } } { \zeta _ \eta}(1 ) { \zeta'_\eta}(5 ) + \frac{{{c_1}}^2 \left ( 14+{c_2 } \right)}{8 { \pi } ^4 } { \zeta'_\eta}(1 ) { \zeta'_\eta}(5 ) \nonumber\\ & & -\frac{9 \left ( 20 + 27 { c_2}+12 { c_4}+{c_6 } \right)}{64 { \pi } ^2 { s_1 } { s_3 } } { { \zeta _ \eta}(3)}^2 -\frac{3 { c_1 } \left ( 3 + 2 { c_2 } \right)}{8 { \pi } ^3 { s_1 } } { \zeta _ \eta}(3 ) { \zeta'_\eta}(3 ) \nonumber\\ & & -\frac{{{c_1}}^2 \left ( 14+{c_2 } \right ) } { 48 { \pi } ^4 } { { \zeta'_\eta}(3)}^2\end{aligned}\ ] ] | the correlation functions of the spin-1/2 @xmath0 chain in the ground state were expressed in the form of multiple integrals for @xmath1 and @xmath2 . in particular , adjacent four - point correlation functions
were given as certain four - dimensional integrals .
we show that these integrals can be reduced to polynomials with respect to specific one - dimensional integrals .
the results give the polynomial representation of the third - neighbor correlation functions . |
according to the standard theory of star formation , stars form out of dense cores embedded in molecular clouds ( e.g. , * ? ? ?
* ; * ? ? ?
the dense cores prior to the protostellar formation , or prestellar cores , are thought to be smooth with no condensations inside @xcite .
the gravitational collapse of such a core impedes the growth of the local density fluctuations , resulting in the formation of at most a couple of condensations , each of which potentially evolves into a protostar that is surrounded by a rotating disk where planets are created .
recent millimeter and submillimeter observations have revealed that the core mass functions ( cmfs ) resemble the stellar initial mass function ( imf ) @xcite .
the resemblance between the cmfs and imf appears to claim that these cores are the direct precursors of individual stars or small stellar systems like binaries .
the above scenario of star formation implicitly assumes that the core has a smooth density distribution and does not contain significant condensations .
this assumption appears to be supported by recent observations by @xcite , who suggested that the prestellar cores do not contain significant substructures on the basis of the interferometric observations with arcsecond resolution ( see also * ? ? ?
@xcite studied substructures around several class 0 sources with subarcsecond resolution and found no clear signs of fragmentation at 100 au . on the other hand
, several previous observations revealed the presence of gravitationally - unbound , substellar - mass structures with sizes of a few thousands au inside prestellar cores in taurus @xcite .
the origins and fates of these small structures inside the prestellar cores remain to be elucidated .
these previous studies of substructures in prestellar cores have concentrated primarily on the relatively - isolated environments .
hence , none of the previous observations have clearly answered the question as to whether prestellar cores in cluster - forming regions have significant substructure or not , because of the lack of high spatial resolution observations [ see , however , @xcite for the discovery of thousand - au scale , substellar - mass structures in oph a ] . however , since most stars form in the clustered environments , it is important to elucidate the structures and properties of prestellar cores in the clustered environments . in the clustered environments ,
the identified cores tend to be more compact than those in quiescent regions , and their sizes are comparable to a telescope beam size . an area covered by a typical single prestellar core in quiescent regions is often occupied by a small cluster of cores in cluster - forming regions @xcite . to verify
whether prestellar cores in the cluster environments contain substantial internal structures or not , we analyze in this letter two representative prestellar cores ( sm1 and b2-n5 ) in @xmath0 ophiuchus , the nearest cluster - forming region at a distance of 120 pc , using the dust continuum data obtained by sma .
the sma data have spatial resolution of a few hundreds au , one of the finest among the available data of prestellar cores .
the first target sm1 , located in the middle of the oph a filamentary cloud , has the strongest submillimeter continuum emission in @xmath0 oph @xcite .
the filamentary morphology is suggestive of interaction with a nearby young b star s1 .
the total mass and dust temperature of sm1 are estimated to be 2 @xmath5 and @xmath6 k , respectively @xcite .
the second target b2-n5 is located in the oph b2 cloud that contains a number of prestellar cores . in the oph b2 cloud ,
a gigantic outflow has recently been detected @xcite .
this core was identified in the n@xmath7h@xmath8 ( 1 - 0 ) emission and is one of the cores having the largest column densities in the oph b2 cloud .
the total mass and temperature are estimated to be about a few tenth @xmath5 ( * ? ? ?
* n@xmath7h@xmath8 ) and @xmath9 k ( * ? ? ?
* nh@xmath10 ) , respectively .
the rest of the paper is organized as follows .
the details of our data are described in section [ sec : data ] .
the results and discussion are presented in sections [ sec : results ] and [ sec : discussion ] , respectively .
the @xmath0 oph cluster - forming region consists of several sub - parsec scale clouds with masses of order of @xmath11 , each of which contains several prestellar cores .
our target cores sm1 and b2-n5 are located in the oph a and b2 clouds , respectively . here
, we briefly describe the dust continuum data toward @xmath0 oph , taken by the single - dish observations .
the 850 @xmath12 m data of the oph a cloud taken with the scuba were obtained through the archival system of the complete survey @xcite .
the fwhm beam size of the jcmt was 14@xmath13 at 850@xmath12 m .
the mean rms noise level of the data is @xmath1410 mjy beam@xmath15 .
the 1.1 mm data of the oph b2 cloud were obtained with the aztec camera @xcite on the aste telescope @xcite .
the fwhm beam size of the aste was 28@xmath13 at 1.1 mm .
the observations were performed in the raster scan mode . each field
was observed several times with azimuth and elevation scans .
the separation among scans was adopted to be 117@xmath13 , which is a quarter of the aztec field of view ( fov @xmath14 7@xmath16 ) .
the effective beam size of the data was 40@xmath13 after the fruit imaging which is an iterative mapping method to recover the spatially - extended component @xcite .
the noise level was less than 10 mjy beam@xmath15 in the entire observed area .
the details of the data will be presented elsewhere . in general , it is difficult to completely recover the emission from extended structures with bolometric observations using the ground - based telescopes due to the atmospheric emission . for scuba ,
the emission from the structures larger than @xmath17 is suppressed in removing the atmospheric emission during the data reduction process , making the 850 @xmath12 m map mostly devoid of the extended emission .
the aztec image also has the same problem for the structures larger than the field of view of @xmath18 .
however , this effect is expected not to be severe because the sizes of the cores and the condensations identified below are smaller than @xmath17 and @xmath19 for scuba and aztec , respectively .
the target prestellar cores sm1 and b2-n5 were observed in the sma compact configuration at 870 @xmath12 m and 1.1 mm , respectively .
we obtained the continuum data through the sma data archive .
the sma is a double - sideband instrument @xcite , having a number of spectral windows of the sma correlator ( `` chunks '' ) .
each target was observed in a single pointing mode .
table [ tab : sma ] summarizes the observational parameters .
the minimum projected baseline lengths were 11 and 12 @xmath20 ( 8.2@xmath13 and 7.5@xmath13 ) at 870 @xmath12 m and 1.1 mm , respectively , and for a gaussian emission distribution with a fwhm of @xmath21 ( @xmath22 800 au ) , the peak flux recovered is @xmath14 50% of the peak flux of the gaussian @xcite .
this problem is called the missing flux problem . to recover the missing fluxes from extended structure and obtain more reliable images and physical properties of the structures , we combined the sma data and the available single - dish data @xcite : i.e. , 850 @xmath12 m scuba data for sm1 and 1.1 mm aztec data for b2-n5 .
below , we mainly discuss the combined sma@xmath23single - dish images .
the combined images are presented in figure [ fig : sma+single ] . for comparison ,
the original sma images are also shown in fig . [
fig : sma ] .
our high spatial dynamic range images constructed by combining the single - dish data and sma data have spatial resolution and sensitivity to dust emission superior to those of existing millimeter and submillimeter observations toward prestellar cores in active cluster - forming regions , without significant effects of the missing flux inherent in any interferometric observations .
both images have clearly revealed the presence of significant substructure inside the prestellar cores . here , we define a compact @xmath24 au - scale structure with a peak intensity larger than 10 @xmath25 rms noise level as a condensation . for sm1
, we identified three condensations , labeled as a1 , a2 , and a3 in fig .
[ fig : sma+single](a ) .
the southernmost condensation , a1 , appears extremely compact , and the internal structure is unresolved even with a few arcsecond resolution .
the existence of the thousands au - scale substructure in the oph a cloud has been pointed out by the previous interferometric observations @xcite .
our image indicates that the substructure previously identified as the strongest 3 mm continuum emission ( core a in figure 2 of @xcite ) is much more compact and consists of two condensations a1 and a2 .
the substructure is more prominent in b2-n5 , which contains several small condensations , as shown in fig .
[ fig : sma+single](b ) . here , we identified 4 condensations in the map .
the separations of the condensations are about 5@xmath2610 arcseconds , or @xmath27 au .
we note that the condensations identified from the combined images can be recognized easily even in the interferometric images alone ( figure [ fig : sma ] ) , and the small - scale structures are essentially the same as those presented in figure [ fig : sma+single ] .
the peak fluxes of all the condensations identified in the sma images alone are more than 4@xmath25 rms noise levels , indicating that these condensations are the real structures existing in the prestellar cores .
table [ tab : condensations ] summarizes the physical properties of the identified condensations .
we applied two - dimensional gaussian fitting to the condensations by using the imfit task in miriad and derived their sizes , positions , and peak and total fluxes .
the masses of the condensations are estimated using the dust opacity of @xmath28 g@xmath15 @xmath29 with @xmath30 @xcite .
this opacity is in agreement with the values adopted by other studies of prestellar cores within a factor of a few ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the typical masses of the condensations are of the order of @xmath31 . for comparison ,
the mass of the critical bonner - ebert sphere whose mean density coincides with the estimated mean density of each condensation is listed in table [ tab : condensations ] .
the elongated beam shapes influence the estimation of the local densities and therefore @xmath32 .
since the identified condensations have sizes comparable to the beam sizes , @xmath32 presented in table [ tab : condensations ] may be underestimated .
almost all the condensations have masses comparable to or larger than the critical bonner - ebert masses ( @xmath32 ) , indicating that the self - gravity plays an important role in the dynamical evolution of the condensations . we note that the masses estimated from the continuum emission depend strongly on the adopted opacity and temperature .
these two parameters are generally difficult to measure accurately .
the dust opacity has uncertainty at least by a factor of a few .
therefore , the masses of the condensations listed in table [ tab : condensations ] are reduced by a factor of a few when the actual dust opacity is smaller .
the adopted temperatures also have uncertainty . here
, @xmath33 k is adopted for sm1 , although a larger temperature of 27 k is reported @xcite .
for b2-n5 , the dust temperature is assumed to be equal to the kinetic temperature of 15 k , which is determined by the nh@xmath10 observations @xcite . in oph b2 ,
the protostellar outflow from el32 @xcite may have directly injected turbulent motions in adjacent dense regions including b2-n5 , and may have increased the temperature of b2-n5 . however , the n@xmath7h@xmath8 ( @xmath34 ) hyperfine fitting indicates that the temperature of b2-n5 appears low @xcite .
the nh@xmath10 observations may not trace the densest gas in oph b2 , and the adopted temperature may be overestimated in the densest parts . even if we take into account these uncertainties of the dust opacity and temperature , all the condensations still have masses that are comparable to or larger than @xmath32 , for sm1 . for oph b2
, the masses of the condensations may become somewhat smaller than @xmath32 when we take into account the larger dust opacity and smaller temperature . even in that case , the existence of the large - amplitude fluctuations at this small - scale is surprising because such large - amplitude fluctuations give a significant impact on the structure formation during the gravitational collapse of the parent core @xcite .
the significant substructure inside the prestellar cores implies that the basic unit of star formation in the clustered environments is much smaller than previously considered .
in fact , recent theoretical and observational studies suggest that prestellar cores formed in the clustered environments tend not to be self - gravitating , but pressure - confined because of the strong inter - core turbulence @xcite . how were these substellar - mass condensations created in parent prestellar cores ?
one possibility is the fragmentation due to internal turbulent motions , or turbulent fragmentation , in short . in the standard theory of star formation
, the cloud turbulence is considered to be less important inside the prestellar cores . however , cluster - forming regions are highly turbulent .
supersonic turbulence potentially triggers the formation of dense regions where the turbulent dissipation preferentially happens @xcite . in such dense regions
the large - amplitude fluctuations can be left . to quantify the dynamical importance of the core internal turbulence , we present in figure [ fig : alpha ] the ratios of the turbulent to gravitational energy ( @xmath35 ) for the prestellar cores in @xmath0 oph as a function of the lte mass , @xmath36 , where @xmath37 is the turbulent energy @xmath38
, @xmath39 is the gravitational energy @xmath40 , @xmath41 is the three - dimensional nonthermal velocity dispersion , @xmath42 is the gravitational constant , and @xmath43 is the core radius .
all the prestellar cores were identified in a high density tracer , h@xmath44co@xmath8 ( @xmath34 ) , using the nobeyama 45 m telescope whose beam size is about 2000 au @xcite .
the mean value of @xmath45 is estimated to be about 0.8 , suggesting that the internal turbulence can contribute significantly to the core dynamics .
this is in contrast to the cores of quiescent regions like taurus @xcite , where @xmath45 tends to be smaller .
in addition , the values of @xmath45 are significantly larger than the threshold for turbulent fragmentation of @xmath46 @xcite .
thus , the turbulent motions are responsible for the formation of the condensations inside the prestellar cores in @xmath0 oph .
h@xmath44co@xmath8 sometimes suffers from the molecular depletion in the dense parts of the prestellar cores .
since @xcite adopted a constant h@xmath44co@xmath8 fractional abundance , the estimated core masses ( and thus @xmath45 ) may be underestimated ( overestimated ) if h@xmath44co@xmath8 is significantly depleted .
however , the depletion effect is expected to be minor as discussed in 4.3 of @xcite .
therefore , we believe that the estimation of the core masses and @xmath45 is reasonable , and our conclusion is not altered by the effects of the molecular depletion . what is the fate of these substellar - mass condensations ? both the prestellar cores contain condensations with sizes of order @xmath47 au and masses of a few hundredth @xmath5 .
almost all the condensations have masses comparable to or larger than the critical bonner - ebert mass .
therefore , they are expected to be gravitationally bound . from the mean separation between the condensations of a few @xmath48 au the coalescence timescale of these condensations is estimated to about @xmath4 yr , comparable to the free - fall times of the individual condensations with densities of @xmath49 @xmath3 , where the relative velocities among condensations are assumed to be equal to the velocity dispersion of b2-n5 , @xmath50 km s@xmath15 .
therefore , these condensations can either merge with themselves or collapse individually to form small groups of low - mass protostars and/or prot - brown dwarfs . here
, the mean separation projected on the plane of the sky is adopted to calculate the coalescence timescale and therefore the estimated coalescence timescale gives a lower limit .
unless the actual mean separation is larger than the adopted value by a factor of @xmath51 , the coalescence timescale is still comparable to the local free - fall time and thus our conclusion is not altered significantly ( see @xcite and @xcite for more quantitative analysis on the separations of condensations . ) .
according to the standard theory of star formation , a typical prestellar core has sizes of 10@xmath52 au and masses of a few @xmath5 @xmath26 10 @xmath5 and is considered to have no significant substructures .
the fragmentation during the prestellar phase is considered to be significantly impeded by the global gravitational collapse before a opaque disk is formed @xcite .
the prestellar cores could fragment into at most a couple of condensations that grow in mass by accretion and eventually evolve into protostars .
thus , a binary or small multiple system is created .
the rotating disks are created around individual protostars and the planets are produced there .
however , the recent hydrodynamic simulation of gravitational collapse of a prestellar core has posed a challenge to this standard picture . in a series of three - dimensional radiation hydrodynamic simulations ,
the protostellar radiation destroys the opaque disks that would evolve into circumstellar disks @xcite .
although whether the disks are completely destroyed or not may depends on the size of the disk formed , this effect of the protostellar radiation significantly constrains the planet formation from the circumstellar disks . in our scenario , substellar - mass condensations
are already formed before the protostars are created .
some of these substellar - mass condensations potentially evolve into giant planets orbiting around protostars or proto - brown dwarfs under the influence of the protostellar radiation that can remove a significant amount of mass from the condensations . here
, we revealed that the two representative prestellar cores in the @xmath0 oph cluster forming region contain hundreds au - scale structures .
another compact structure with @xmath141000 au was also found toward the southern part of oph a , oph a - n6 @xcite .
very recently , @xcite discovered a surprising number of brown dwarf candidates in this region .
although the follow - up spectroscopic observations would be needed , they estimated the ratio of the brown dwarfs ( @xmath53 ) to low - mass stars ( @xmath54 ) to be about 1.4 , more than 5 times larger than the previous estimation .
our observations may explain the abundance of substellar - mass objects like brown dwarfs in this region .
however , it remains unclear whether the small - scale structures are really common in prestellar cores in the clustered environments .
further observations with higher spatial resolution and higher sensitivity will be needed to clarify the internal structures of the prestellar cores .
the high - angular resolution interferometric observations such as alma will provide us an excellent opportunity to explore the substructures of prestellar cores .
we thank the referee for valuable comments that improve the paper .
this work was supported partly by a grant - in - aid for scientific research of japan , 20540228 and 22340040 ( f. n. ) and a grant from the national science council of taiwan nsc 99 - 2112-m-001 - 013-my3 ( s. t. ) .
99 andr , p. , ward - thompson , d. , & barsony , m. , 1993 , , 406 , 122 andr , p. , ward - thompson , d. , & motte , f. , 1996 , , 314 , 625 bacmann , a. , andr , p. , puget , j .- l . , abergel , a. , bontemps , s. , & ward - thompson , d. 2000 , , 361 , 555 barsony , m. , haisch jr . ,
k. e. , march , k. a. , & mccarthy , c. , 2012 , , 751 , 22 bourke , t. l. , myers , p. c. , caselli , p. et al . , 2012 , , 745 , 117 ezawa , h. , kawabe , r. , kohno , k. , & yamamoto , s. 2004 , _ proc .
spie _ , 5489 , 763 evans ii , n. j. , dunham , m. m. , jorgensen , j. k. et al .
2009 , , 181 , 321 friesen , r. k. , di francesco , j. , shimajiri , y. , & takakuwa , s. , 2010 , , 708 , 1002 friesen , r. k. , di francesco , j. , shimajiri , y. , & takakuwa , s. , 2010 , , 708 , 1002 goodwin , s. p. , whitworth , a. p. , & ward - thompson , d. , 2004 , , 414 , 633 hennebelle , p. & teyssier , r. , 2008 , , 477 , 25 hildebrand , r. h. , 1983 , q. j. r. astron .
24 , 267 ho , p. t. p. , moran , j. m. , & lo , k. y. , 2004 , , 616 , l1 johnstone , d. , wilson , c. d. , moriarty - schieven , g. , et al . , 2000 , , 545 , 327 kamazaki , t. , saito , m. , hirano , n. , & kawabe , r. , 2001 , , 548 , 278 langer , w. d. , velusamy , t. , kuiper , t. b. h. , levin , s. , olsen , e. , & migenes , v. 1995 , , 453 , 293 liu , g. , calzetti , d. , yun , m. s. , et al .
2010 , , 139 , 1190 machida , m. n. , tomisaka , k. , matsumoto , t. , & inutsuka , s. , 2008 , , 677 , 327 maruta , h. , nakamura , f. , nishi , r. , ikeda , n. , & kitamura , y. , 2010 , , 714 , 680 maury , a. j. , andr , p. hennebelle , p. et al .
2010 , , 512 , 40 mckee , c. f. & ostriker , e. c. , 2007 , , 45 , 565 motte , f. , andr , p. , & neri , r. , 1998 , , 336 , 150 nakamura , f. , kamada , y. , kamazaki , t. , et al .
, 2011 , , 726 , 46 nakamura , f. & li , z .- y . , 2011 , , 740 , 36 olmi , l. , testi , l. , & sargent , a. i. 2005 , , 431 , 253 ossenkopf , v. & henning , t. 1994 , , 291 , 943 peng , r. , langer , w. d. , velusamy , t. , kuiper , t. b. h. , & levin , s. 1998 , , 497 , 842 ridge , n. a. , di francesco , j. , kirk , h. , et al . 2006 , , 131 , 2921 roy , n. , datta , a. , momjian , e. , & sarma , a. p. 2011
, , 739 , l4 , 131 , 2921 schnee , s. , enoch , m. , johnstone , d. , et al . , 2010 , , 718 , 306 schnee , s. , sadavoy , s. , di francesco , j. & johnstone , d. , 2012 , , 755 , 178 schonke , j. & tscharnuter , w. m. , 2011 , , 526 , 139 shu , f. h. , adams , f. c. , & lizano , s. 1987 , , 25 , 23 stanke , t. , smith , m. d. , gredel , r. , & khanzadyan , t. , 2006 , , 447 , 609 takakuwa , s. , kamazaki , t. , saito , m. , & hirano , n. 2003 , , 584 , 818 takakuwa , s. , ohashi , n. , bourke , t. l. , et al .
, 2007 , , 662 , 431 wilner , d. j. , & welch , w. j. 1994 , , 427 , 898 wilson , g. w. , austermann , j. e. , perera , t. a. , et al . , 2008 , , 386 , 807 lll observing date & july 29 , 2007 & june 1 , 2008 + number of antennas & 7 & 8 + r.a .
( j2000 ) & 16:26:27.60 & 16:27:29.15 + decl.(j2000 ) & -24:23:55.0 & -24:27:02.0 + primary beam hpbw & 36@xmath13 & 45@xmath13 + synthesized beam hpbw & @xmath55 ( p.a.=49@xmath56 ) & @xmath57 ( p.a.=40@xmath56 ) + baseline coverage & 10.9 - 148.2 ( k@xmath58 ) & 12.1 - 114.2 ( k@xmath58 ) + lo frequency & 341.5 ghz & 274.2 ghz + bandwidth & 4.0 ghz & 3.8 ghz + gain calibrator & nrao530 & 1625 - 254 + flux of the gain calibrator & 1.4 jy & 0.79 jy + passband calibrator & 3c273 , 3c454.3 & 3c279 , jupiter + flux calibrator & uranus & titan + system temperature & 150 - 600 ( k ) & 300 - 600 ( k ) + rms noise level & 4.8 mjy / beam & 1.2 mjy / beam [ tab : sma ] lllllllll a1 & 16:26:27.8 & -24:24:00 & 455 & 742 & @xmath59 & 0.15 & 30.7 & 0.014 + a2 & 16:26:27.7 & -24:23:57 & 176 & 958 & @xmath60 & 0.17 & 4.6 & 0.036 + a3 & 16:26:26.6 & -24:23:54 & 138 & 769 & @xmath61 & 0.13 & 2.2 & 0.053 + b1 & 16:27:29.6 & -24:27:07 & 15.2 & 44 & @xmath62 & 0.033 & 1.4 & 0.03 + b2 & 16:27:29.5 & -24:27:04 & 17.4 & 136 & @xmath63 & 0.10 & 0.53 & 0.056 + b3 & 16:27:28.9 & -24:27:02 & 15.8 & 47 & @xmath64 & 0.035 & 1.3 & 0.036 + b4 & 16:27:28.6 & -24:26:59 & 13.1 & 49 & @xmath65 & 0.036 & 0.79 & 0.046 [ tab : condensations ] | we present combined submillimeter - array ( sma ) + single - dish images of the ( sub)millimeter dust continuum emission toward two prestellar cores sm1 and b2-n5 in the nearest star cluster forming region , @xmath0 ophiuchus .
our combined images indicate that sm1 and b2-n5 consist of three and four condensations , respectively , with masses of @xmath1 and sizes of a few hundred au .
the individual condensations have mean densities of @xmath2 @xmath3 and the masses are comparable to or larger than the critical bonner - ebert mass , indicating that the self - gravity plays an important role in the dynamical evolution of the condensations .
the coalescence timescale of these condensations is estimated to be about @xmath4 yr , which is comparable to the local gravitational collapse timescale , suggesting that merging of the condensations , instead of accretion , plays an essential role in the star formation process .
these results challenge the standard theory of star formation , where a single , rather featureless prestellar core collapses to form at most a couple of condensations , each of which potentially evolves into a protostar that is surrounded by a rotating disk where planets are created . |
we are interested here in estimating the error term @xmath6 associated with the van der corput transform , @xmath7 where @xmath2 and @xmath1 are several times continuously differentiable functions with @xmath8 for @xmath9 $ ] and where @xmath10 is defined by @xmath11 .
a starred sum indicates that if a limit of summation is an integer , the corresponding summand is multiplied by @xmath12 .
the function @xmath13 denotes @xmath14 .
the van der corput transform is best - known in number theory for being the crucial element of process b in the theory of exponent pairs and is sometimes simply referred to as `` process b. '' as a method of estimating exponential sums , the van der corput transform is often presented alongside other methods of weyl , van der corput , and vinogradov .
direct application of the van der corput transform can take a complicated sum to one more amenable to estimation techniques or it can reduce the number of terms and therefore make computational estimations easier .
the van der corput transform itself is involutive applying it to the right - hand side of will simply return the sum on the left - hand side of so one gains nothing by applying it twice in a row ; but one could alternate applications of the van der corput transform with other techniques ( such as the process a of the theory of exponent pairs ) to achieve better results .
this alternation method is still a fruitful ground for modern research .
recently , cellarosi @xcite attained interesting new results in the classical case of @xmath4 and @xmath15 , where the alternating technique employed is simply reducing @xmath16 modulo @xmath17 ; nakai @xcite has investigated the possibility of using an analogous method when @xmath18 is cubic or quartic .
the van der corput transform has also seen recent use in physical applications , including quantum optics and wave processes ( see @xcite and the papers cited there for more details ) .
van der corput @xcite originally showed that , given @xmath19,\ ] ] the error term can be bounded like @xmath20 ( here we use the landau and vinodradov asymptotic notations which will be defined explicitly in section [ section : notation ] . )
phillips @xcite improved this error under the additional assumptions @xmath21
\qquad \text{and } \qquad \lambda_3 ^ 2=\lambda_2 \lambda_4;\ ] ] in this case , we can replace @xmath22 with @xmath23 .
the form of the error term found in most modern texts on analytic number theory @xcite has its roots in the work of kolesnik @xcite and heath - brown @xcite , although the results of the latter authors required analyticity of the function @xmath2 , an assumption which has since been circumvented .
this moderately - difficult - to - prove form of the error term suffices for many basic application of the van der corput transform .
we present this modern bound on the error in the notation used by huxley .
[ thm : textbook](lemma 5.5.3 in @xcite ) suppose that @xmath18 is real and four times continuously differentiable on @xmath3 $ ] .
suppose that there are positive parameters @xmath24 and @xmath25 , with @xmath26 , such that , for @xmath9 $ ] , we have @xmath27 let @xmath28 be a real function of bounded variation @xmath29 on the closed interval @xmath3 $ ] .
then @xmath30 where @xmath10 is the unique solution in @xmath3 $ ] to @xmath11 .
the implicit constant in the big - o term depends on the implicit constants in the relations between @xmath25 , @xmath24 , and the derivatives of @xmath18 .
the error term @xmath31 here corresponds to the @xmath32 term in the estimates of van der corput and phillips .
unfortunately , for many interesting cases , the above error is insufficient . as huxley @xcite notes , when applying the van der corput transform to a multi - dimensional exponential sum , `` ... the error terms and the truncation error in the poisson summation formula may add up to more than the estimate for the reflected sum . ''
thus , finer error terms , useful for a broad spectrum of problems including computation and physical applications , have been given by various people , including kolesnik @xcite , liu @xcite , redouaby and sargos @xcite , and karatsuba and korolev @xcite .
liu extends an ( unfortunately obscure ) earlier work of min @xcite , removing the latter s condition that @xmath18 be an algebraic function .
redouaby and sargos show that the conditions on @xmath33 and @xmath34 can be removed without greatly increasing the bound on the error term .
the work of karatsuba and korolev is unique among all papers cited here , in that it is the only to give the implicit constants in the big - o terms explicitly , making it superior for most computational work . outside of redouaby and sargos result
, we can not briefly state any of these other forms of the error term in full detail ; we will , however provide the following inexplicit form of karatsuba and korolev s result as an example of the comparative strength of these errors compared with theorem [ thm : textbook ] .
[ thm : karatsubakorolev ] suppose that @xmath18 and @xmath28 are real - valued functions with @xmath2 four times continuously differentiable and @xmath1 two times continuously differentiable on the interval @xmath3 $ ] .
suppose there are positive constants @xmath24 , @xmath25 , and @xmath35 , with @xmath36 , such that , for @xmath37 $ ] , @xmath38 where @xmath10 is the unique solution to @xmath11 in the interval @xmath3 $ ] and @xmath39 the size of the implicit constant in the big - o term depends on the implicit constants in the relations of @xmath24 , @xmath25 , @xmath35 , and the derivatives of @xmath18 and @xmath28 . if , in addition , @xmath40 and @xmath41 , we may remove the terms @xmath42 .
the @xmath31 term again makes an appearance in this theorem .
it can not , in fact , be completely removed .
if @xmath43 is not an integer , but @xmath44 is an integer , then changing @xmath43 to @xmath45 for some very small @xmath46 will not change the value of the sum on the left - hand side of , but removes a term of size @xmath31 from the right - hand side .
van der corput s results in @xcite also deserve a mention here , as they are of an entirely different flavor from those listed above and because they are generally not very well - known . instead of having a coefficient of @xmath47 in the transformed sum , he has a more general set of coefficients .
in addition , his error term which would require too many definitions to state succinctly here bears no resemblance to any of the other error estimates cited or formulated in this paper . while van der corput s results in @xcite are quite complicated to use , they are aesthetically pleasing .
as we remarked above , the van der corput transform is involutive , but in all the other results given above as well as the main result of this paper , one can obtain very different error terms when one applies the transform to the right - hand side of instead of the left , assuming that the conditions necessary to apply the results would even still hold . in @xcite ,
van der corput shows that the conditions are still satisfied and the error term unchanged regardless of which side one applies his transform to ; his _ theorem _ is involutive .
the van der corput transform and its error has been studied in much more general settings than we go into here : of particular interest , jutila @xcite considered sums of the form @xmath48 for certain multiplicative functions @xmath49 ( see also ( * ? ? ?
* ch . 20 ) ) , and krtzel @xcite considered the van der corput transform of a convergent infinite series .
one may ask what the best possible error term could be . given the frequent restriction in theorems on the van der corput
transform that @xmath50 ( or , equivalently @xmath51 ) , it is not surprising that the case where @xmath1 is constant and @xmath2 is quadratic ( that is , @xmath52 ) is the most - commonly studied special case of the van der corput transform and the one with the best error terms @xcite .
fedotov and klopp @xcite that helped to spark the author s investigation into the van der corput transform . ]
have given the error term in this case as an explicit integral .
but perhaps most amazing are the results of coutsias and kazarinoff @xcite : they showed that for positive integers @xmath53 , we have @xmath54 for @xmath55 , @xmath56 is the nearest integer to @xmath57 , and @xmath58 is a particular constant . not only is the error bounded but it shrinks to zero as @xmath57 nears an integer .
the coutsias - kazarinoff result suggests that the van der corput transform should be very accurate ; in particular , the van der corput transform for nice enough functions @xmath2 and @xmath1 should nt have compounding error terms ( such as the @xmath59 ) seen in all the other results mentioned above . the main theorems of this paper ( and theorem [ thm : toinfinity ] , especially ) confirm this hypothesis , allowing the van der corput transform to be applied on very long intervals with a much higher degree of accuracy than in previous results .
as a quick example of this , consider the following well - known transform which appears in iwaniec and kowalski s book @xcite ( a version of this transform also appears in @xcite ) .
given @xmath60 , @xmath61 , and @xmath62 , @xmath63 , consider @xmath64 where @xmath65 , @xmath66 , and @xmath67
. using a form of theorem [ thm : textbook ] , iwaniec and kowalski give @xmath68 with an implicit constant dependent on @xmath16 and @xmath69 . using the results of this paper , we may improve this to the following .
[ cor : iwakowimprovement ] provided @xmath70 and @xmath71 , in equation , we may take @xmath72 with an implicit constant dependent _ only _ on @xmath16 and the implicit constants in the lower bounds on @xmath73 and @xmath24 .
corollary [ cor : iwakowimprovement ] allows us to take @xmath69 as large as we like ( and hence our intervals as large as we like ) without increasing the bound on @xmath6 .
this , as well as the loss of the logarithmic factor , are common to applications of the results in this paper .
more powerful results are possible . in many cases ,
the main theorems of this paper also allow one to extract the next term in the asymptotic for the van der corput transform .
under the same general hypotheses of theorems [ thm : textbook ] and [ thm : karatsubakorolev ] , we can improve the error terms to the following .
[ thm : karatsubakorolevimprovement ] suppose that @xmath18 and @xmath28 are real - valued functions with @xmath2 four times continuously differentiable and @xmath1 two times continuously differentiable on the interval @xmath3 $ ] .
suppose there exist constants @xmath24 , @xmath25 , and @xmath35 satisfying @xmath74 , @xmath75 , and the bounds on lines and .
then , @xmath76 where @xmath77 equals @xmath78 here , @xmath79 represents the nearest integer for an integer @xmath53 , then it does nt matter if we let @xmath79 equal @xmath53 or @xmath80 provided that we do so consistently . ]
to @xmath81 ; @xmath82 , the difference between @xmath81 and the nearest integer to @xmath81 , namely @xmath83 ; and @xmath84 , the distance between @xmath81 and the nearest integer to @xmath81 , so that @xmath85 .
the function @xmath86 equals @xmath87 which converges and is uniformly bounded for all real @xmath81 .
the size of the implicit constant in the big - o term depends on the implicit constants in the relations of @xmath24 , @xmath25 , @xmath35 , and the derivatives of @xmath18 and @xmath28 .
this in particular suggests that the size of @xmath88 in karatsuba and korolev s result is optimal when @xmath89 .
when @xmath90that is , when @xmath91 is an integer
the term @xmath77 can bemore simply bounded by @xmath92 .
corollary [ thm : karatsubakorolevimprovement ] can , in certain cases , be used to improve the kusmin - landau inequality , a common result in the study of exponential sums .
( the kusmin - landau inequality theorem 2.1 in @xcite ) suppose @xmath18 is continuously differentiable and that @xmath93 is monotonic on an interval @xmath3 $ ] .
moreover suppose @xmath94 on @xmath3 $ ] , where @xmath95 is the distance from @xmath93 to the nearest integer .
then @xmath96 [ thm : expk - l ] suppose @xmath2 , @xmath25 , and @xmath24 satisfy the conditions of corollary [ thm : karatsubakorolevimprovement ] .
if @xmath97}\|f'(z)\|$ ] is positive , @xmath98 , and @xmath99 , then @xmath100 where @xmath101 represents the difference between @xmath81 and the nearest integer to @xmath81 . corollary [ thm : expk - l ] follows immediately as a special case of corollary [ thm : karatsubakorolevimprovement ] .
moreover , it suggests that the constant in the classical kusmin - landau inequality may not be optimal .
the function @xmath102 grows like @xmath103 as @xmath104 approaches zero ; however , corollary [ thm : expk - l ] suggests that the growth should be at worst like @xmath105 as @xmath104 approaches zero .
while corollaries [ thm : karatsubakorolevimprovement ] and [ thm : expk - l ] strengthen many previous results , they are constrained to apply to short intervals .
the results of this paper do not give similarly simple conditions and error terms when the size of the interval is allowed to grow arbitrarily large .
but , given a specific sum , we can show very great improvements as one endpoint of the interval tends towards infinity .
consider the specific transform @xmath106 where @xmath73 is an integer .
( we have kept the @xmath107 due to it s natural appearance in the application of the van der corput transform .
it may be replaced by @xmath12 with no change in value on the right - hand side , however . )
if we apply the form of the error term from theorem [ thm : textbook ] , then we obtain the following result , which was included as an example in @xcite .
[ cor : monterror ] in line , we have @xmath108 if we apply theorem [ thm : karatsubakorolev ] from karatsuba and korolev instead , we obtain much finer results .
[ cor : karakorerror ] in line , we have @xmath109 where @xmath84 is the distance from @xmath81 to the nearest integer . in this case
, the @xmath110 term comes from needing to break the sum on the left - hand side of into roughly diadic intervals , each of which contributes an error term of size around @xmath111 .
one must also choose the endpoints of these intervals to be values which are @xmath112 times a square , in order to keep the error term @xmath88 zero . using the results of this paper
, we can show that the result of karatsuba and korolev is almost sharp , in that we will extract an explicit term of size @xmath113 when @xmath113 is smaller than @xmath114 and extract an explicit term of size @xmath114 when @xmath114 is smaller than @xmath113 .
[ thm : exampleerror ] in line , we have @xmath115 where @xmath116 is a particular constant give @xmath117 . ] and @xmath118 is the sawtooth function .
the remaining functions are all as in corollary [ thm : karatsubakorolevimprovement ] .
one way to understand these new error terms is via the geometry of the associated sums . given functions @xmath2 and @xmath1 , one typically considers the curve @xmath119 given by @xmath120 where @xmath121 represents the floor of a real number @xmath81 and @xmath122 represents the fractional part of @xmath81 .
geometric aspects of these curves have been well - studied @xcite .
especially when @xmath123 is small , these curves generate a series of spiral - like figures .
the van der corput transform of the sum associated to @xmath124 then generates a new curve that can be seen to connect the center point of successive spirals by straight lines like in figure [ fig : nlogn ] .
thus the van der corput transform can be seen as smoothing ( if @xmath123 is small ) or roughening ( if @xmath123 is big ) the curve .
an easily accessible explanation for why this occurs is given in @xcite .
for our example case , the corresponding curve is displayed in figure [ fig : exampleerror ] .
the term @xmath125 from corollary [ thm : exampleerror ] describes the outer arm of the spirals and the connections between them .
figure [ fig : outerarm ] depicts the outer arm of the @xmath126th spiral and its approximation using corollary [ thm : exampleerror ] .
the reason why this approximation appears so poor is due , in part , to the number of points on this part of the spiral .
the @xmath127th spiral has on the order of @xmath127 points in it .
the outer arm of the spiral , where @xmath128 accounts for only about @xmath129 points .
the big - o term @xmath130 is on the order of the explicit term when @xmath131 , so that is a good estimate for only @xmath132 points , a negligible piece of the spiral .
however , from the picture , we can see that the approximation does a very good job of estimating the rotation around the spiral and appears to be of the correct order of magnitude for the distance to the center of the spiral .
although we do not calculate the big - o constants explicitly , it is quite possible the big - o terms are smaller in modulus than for @xmath133 .
it is possible to obtain far more explicit results along the outer spiral by making use of the fresnel integral functions ; however , this method does not provide good asymptotic data since it uses values of the fresnel integral functions for which good asymptotics do not currently exist .
more details will be given in section [ section : fresnel ] .
the remaining points ( in fact , almost all of the points ) are described by the term @xmath134 which gives the tight inner weave of a given spiral .
the term @xmath135 can be bounded by @xmath136 for all @xmath73 .
the big - o term @xmath137 is bigger than @xmath136 for only @xmath138 of the points on the @xmath127th spiral , again a negligible piece .
thus , we know almost all points on the inner weave ( and hence almost all points on the full curve ) to within a @xmath136 error and often much better .
the big - o terms grow to the size of the explicit terms and as @xmath139 approaches @xmath140 in both cases , so we do nt have good information about how the curve transitions from the inner weave to the outer arm and vice - versa .
we also do not have a good heuristic explanation for why @xmath141 appears in on the outer arm , but @xmath142 appears in for the inner weave .
this paper is arranged as follows . in section [ section : notation ] , we will provide a short catalog of notations that are frequently used in this paper . in section [ section : heuristics ]
, we will outline the methods of this paper and how they overcome difficulties encountered in other results . in section [ section :
results ] , we will state the main theorems of this paper as well as a briefly discuss them and their assumptions ; in this section , we also include a proof of corollaries [ cor : iwakowimprovement ] and [ thm : karatsubakorolevimprovement ] . in section [ sec : lemstationaryphase ]
, we will cite a number of lemmas from other papers and books that will be necessary in the proof of the theorems , as well as give propositions that show how the lemmas apply under the specific conditions of the main theorems . in the remaining sections ,
we prove the primary theorems of our paper and then prove corollary [ thm : exampleerror ] .
we will frequently use the landau and vinogradov asymptotic notations .
the big - o notation @xmath143 ( equivalently , @xmath144 ) means that there exists some constant @xmath116 such that @xmath145 on the domain in question . by @xmath146 ,
we mean that a function which is asymptotically bounded by @xmath18 will also be asymptotically bounded by @xmath28 .
the little - o notation @xmath147 as @xmath148typically with @xmath149 equal to @xmath150means that @xmath151 constants in the subscript of big - o notation , such as @xmath152 , mean that the implicit constant @xmath116 may depend on the subscript constants ( in this case , @xmath153 ) . by @xmath154 , we shall mean that @xmath155 .
we will need several functions that relate a real number @xmath81 to the integers nearby it .
let @xmath156 denote the usual floor of a real number @xmath81 , the largest integer less than or equal to @xmath81 , and let @xmath157 denote the usual ceiling of a real number @xmath81 , the smallest integer greater than or equal to @xmath81 .
let @xmath158 denote the fractional part of a real number @xmath81 .
let @xmath159 denote the sawtooth function , and let @xmath160 denote the `` smoothed '' sawtooth function , with fourier series @xmath161 we also wish to have a modified sawtooth function , given by @xmath162 the convergence of @xmath163 will be guaranteed by proposition [ prop : fouriertail ] .
let @xmath164 denote the difference between @xmath81 and the nearest integer to @xmath81 , let @xmath165 denote the nearest integer to @xmath81 , and let @xmath166 denote the distance from @xmath81 to @xmath167 .
let @xmath168 be given by @xmath169 a few more definitions will simplify the ( nonetheless still complicated ) statement of the theorem . for visibility s sake
, we will avoid writing @xmath170 all the time when the choice of argument is always the same .
given functions @xmath2 and @xmath1 , let @xmath171 and given an integer @xmath172 , let @xmath173
the transform is a simple two - step process : the poisson summation formula is first applied to obtain a sum of integrals , and then each integral is estimated using stationary phase methods . in particular , the poisson summation formula ( see lemma [ lem : poissonsummation ] ) gives @xmath174 an oscillatory integral , like @xmath175 is said to have a stationary phase point if there exists an @xmath176 $ ] such that @xmath177 ; if such a point exists , then we expect to be roughly @xmath178 plus some small error , where @xmath179 equals @xmath12 if @xmath180 equals @xmath149 or @xmath43 and @xmath17 otherwise . in our case ,
the stationary phase point of the integrand @xmath181 occurs at @xmath182 , which is defined by @xmath183 .
provided @xmath184 , this stationary phase point will be inside the interval of integration , and so we expect @xmath185 the error term for the van der corput transform thus comes from the error in estimating all the integrals of .
the first problem in doing such an estimation comes from the integrals without a stationary phase point .
typically the method one uses to estimate such an integral is an application of integration by parts .
@xmath186_a^b + \int_a^b \frac{f''(x)}{2\pi i(f'(x)-r)^2 } e(f(x)-rx ) \ dx\ ] ] ( we will assume that @xmath4 for a while to simplify the heuristics . ) the terms @xmath187 at @xmath149 and @xmath43 are referred to as the first - order endpoint contributions and exhibit a great deal of cancellation ( see proposition [ prop : fouriertail ] ) ; for example , if @xmath188 is an integer , then @xmath189 and similarly if @xmath44 is an integer .
however , we are still left with the integral @xmath190 to evaluate .
we could estimate the integral by taking absolute values of the integrand , but we would then obtain an estimate of the size @xmath191 , which is of the same order as the first - order endpoint contributions , but without the cancellation that would allow them to be nicely summed .
one possible solution would be to apply integration by parts a second time : this gives @xmath192_a^b\\ - \int_a^b \frac{d}{dx}\left ( \frac{f''(x)}{(2\pi i)^2(f'(x)-r)^3 } \right ) e(f(x)-rx ) \ dx,\end{gathered}\ ] ] with a second - order endpoint contribution and a new integral .
the second - order endpoint terms are absolutely convergent , and if @xmath188 and @xmath44 are integers , they sum to @xmath193 , which is a good estimate if @xmath123 is small at the endpoints . on the other hand ,
if we try to bound the new integral in by taking absolute values of the integrand , we get terms of the same order of magnitude as the second - order endpoint contributions ( which , unlike the first - order endpoints , are summable ) . but in addition , we have the total variation of @xmath194 on @xmath3 $ ] , which may be roughly bounded by the sum of the moduli of local maxima and minima of @xmath194 on @xmath3 $ ] . unless @xmath18 is quadratic as in coutsias and kazarinoff s case it is difficult to find good estimates on these integrals .
more applications of integration by parts or the presence of a non - constant @xmath1 only make things worse .
therefore , these terms are often avoided entirely by applying the truncated poisson formula instead of the full poisson summation formula .
( truncated poisson formula
proposition 8.7 in @xcite ) is given in lemma 7 of @xcite . ]
let @xmath18 be a real function with @xmath8 on the interval @xmath3 $ ] .
we then have @xmath195 where @xmath16 , @xmath196 , and @xmath197 are any numbers with @xmath198 and @xmath199 , the implied constant being absolute .
the truncated poisson formula is the source of the @xmath200 error term in many of the results mentioned above . after applying the truncated poisson formula with @xmath201 and @xmath202 ,
one is left with approximately @xmath203 integrals with which one hopes to apply stationary phase estimates .
these estimates work best on a small interval around the stationary phase point , where the second derivative of @xmath2 is fairly constant , and the higher derivatives of @xmath2 are small .
so we would like to break the integrals in into several pieces , such as @xmath204 where the point of stationary phase is near the middle of @xmath205 $ ] . however , to make this effective , we would again require good estimates on integrals with no stationary phase point ( the integrals from @xmath149 to @xmath206 and @xmath207 to @xmath43 ) .
in addition , one is faced with possible first - order ( and higher ) endpoint contributions at @xmath206 and @xmath207 .
( while estimates of stationary phase integrals benefit from integrating on a small interval , they suffer again if the interval is too small .
the @xmath208 terms on line that appear in stronger results on the van der corput transform arise from bounding the stationary phase integrals where @xmath10 is closest to @xmath149 or @xmath43 . ) because of this , many results on the van der corput transform seek to treat the entire integral from @xmath149 to @xmath43 as a single stationary phase integral , hence the common conditions , as in theorem [ thm : textbook ] , that @xmath209 has constant order on the entire interval @xmath3 $ ] and that higher derivatives of @xmath2 be likewise small on the entire interval @xmath3 $ ] .
the techniques of this paper seek to overcome some of these difficulties .
first , we use the recent and powerful stationary phase estimates first proved by huxley @xcite and refined by redouaby and sargos @xcite .
these estimates show that stationary phase integrals , such as the middle term in , also generate first - order endpoint contributions .
in fact , these contributions directly cancel the first - order endpoint contributions at @xmath206 and @xmath207 generated by the first and third term in .
thus we no longer need to treat the full integral from @xmath149 to @xmath43 with stationary phase estimates and so can replace the global restrictions on the derivatives of @xmath2 and @xmath1 ( such as those in theorem [ thm : karatsubakorolev ] ) with local restrictions .
in particular , we replace the the standard assumptions @xmath210 $ ] and @xmath37 $ ] with some function @xmath211 ( see section [ section : conditionm ] for more details about the function @xmath211 ) .
second , to avoid use of the truncated poisson summation formula , we develop a method to get reasonable bounds on the integrals arising from applying integration by parts twice , as in . instead of looking at one integer @xmath172 at a time and counting the contribution of @xmath194 for each @xmath81 s which give local maxima and minima , we instead look at each @xmath81 and imagine a possible fixed real - valued @xmath172 that causes the point @xmath81 to be a critical point of this function .
this then defines a function @xmath212 , and we estimate the sum over all values of @xmath81 where the function @xmath212 takes integer values using a variant of euler - maclaurin summation ( see proposition [ prop : inversesum ] and section [ sec : variation ] ) .
third , since the previous two techniques frequently benefit from @xmath123 being large , we develop a method to deal with the large second - order endpoint contributions in this case .
the reason for these terms being large is that @xmath213 is large when @xmath172 is close to @xmath188 ( and likewise at @xmath43 ) .
however , if @xmath214 is large and @xmath215 not too large in comparison , then a small shift in @xmath149 to , say , @xmath216 should greatly increase the size of the denominator while keeping the numerator roughly the same size . therefore , by altering the endpoints of the integrals in the poisson summation formula by a small amount when @xmath172 is close to @xmath188 or @xmath44
, we can reduce the size of the second - order endpoint contributions . with these techniques , the main theorems of this paper give the van der corput
transform in the following form .
@xmath217 the @xmath218 terms will usually equal the first - order endpoint contributions and will be explicit in many cases .
the term @xmath219 will estimate the error that arises when there exists an @xmath10 close to @xmath149 or @xmath43 and is analogous to the @xmath220 terms present in the results of kolesnik , liu , and karatsuba and korolev . the term @xmath221 will estimate the second - order endpoint contributions at @xmath149 and @xmath43 .
the term @xmath222 ( and a bit of @xmath223 ) will estimate the remaining error in applying stationary - phase methods and the second - order endpoint contributions at @xmath206 and @xmath207 .
the remainder of @xmath223 will estimate the size of the integrals that arise from applying integration by parts twice .
the techniques outlined in this section allow for a number of interesting extensions to the van der corput transform .
since we replace the global restrictions on the derivatives of @xmath2 and @xmath1 with local restrictions , we can apply the van der corput transform to many new sums , including those where @xmath2 and @xmath1 have moderate oscillations .
likewise , we can now directly apply the van der corput transform to sums where @xmath123 is large . the function @xmath224 that appears on the right - hand side of the van der corput transform is sometimes referred to as the van der corput reciprocal of the function @xmath2 .
is perturbed . ]
if the second derivative of @xmath2 is large , then the second derivative of the van der corput reciprocal of @xmath2 will be small , so the results cited above , if they can not be applied directly to a sum , can be and frequently are applied `` backwards '' to the van der corput transform of the sum . for example
, if we attempted to apply the form of the error from theorem [ thm : karatsubakorolev ] directly to the sum @xmath225 for some @xmath226 , then the error term @xmath200 would be at least @xmath227 , the size of the trivial bound on the sum ( and that does nt even take into account the size of the implicit constant ! ) .
theorem [ thm : main ] allows us to apply the transform to this sum directly and obtain a @xmath136 error .
in general , if @xmath228 is reasonably large compared to @xmath1 and both are free of wild oscillations , then @xmath6 should be no larger than @xmath229 . as a final note
, the results of this paper do not give an improvement in all possible cases .
if , for example , @xmath123 varies between a very large value and a value very close to zero , then our results may give very poor error terms .
another case where our results do not give any improvement is @xmath230
we will use a function @xmath211 to measure the length of an interval around @xmath81 where the functions @xmath209 and @xmath28 are well - approximated by their taylor polynomials up to the second degree .
the larger @xmath211 is , the more linear the functions @xmath209 and @xmath28 will appear at the point @xmath81 . to be more concrete , by condition @xmath231 , we shall refer to the existence of positive , bounded , continuously differentiable functions @xmath211 and @xmath232 on @xmath3 $ ] , along with several associated positive constants @xmath233 , @xmath234 , @xmath235 , @xmath236 , @xmath237 , @xmath238 , @xmath239 , which together satisfy several conditions : 1
. @xmath240 ; 2 .
@xmath241 , and let @xmath242 satisfy @xmath243 ; 3 . if @xmath244 $ ] , where @xmath245 then @xmath2 is @xmath246 $ ] and @xmath1 is @xmath247 $ ] ; and , 4 .
if @xmath248 denotes the intersection @xmath249\cap j$ ] , then for all @xmath9 $ ] we expect all the following inequalities to hold for @xmath250 : @xmath251 for any given @xmath18 , @xmath28 , @xmath149 , and @xmath43 , there are infinitely many possible choices of the functions @xmath211 , @xmath232 , and the associated constants that satisfy the above conditions . therefore , for the remainder of the paper , we shall assume that if @xmath18 , @xmath28 , @xmath149 , and @xmath43 remain unchanged , the particular functions @xmath211 and @xmath232 and associated constants we reference will be likewise unchanged . if @xmath18 and @xmath28 are unchanged , but @xmath149 and @xmath43 allowed to vary , then we will assume , first , that the associated constants will be unchanged , and , second , that given two intervals @xmath252 $ ] and @xmath253 $ ] with corresponding auxiliary functions @xmath254 , @xmath255 and @xmath256 , @xmath257 , we have @xmath258 and @xmath259 on @xmath260 .
since condition @xmath231 is rather intricate , we pause a moment to illuminate it further . 1
. the requirement of @xmath241 and @xmath243 in condition @xmath231 part ( ii ) guarantee that the taylor approximations to @xmath123 has good properties ( see proposition [ prop : fbound ] , lemma [ lem : redsar1 ] , and the remark following the lemma ) .
the restriction of condition @xmath231 part ( iii ) to have @xmath2 and @xmath1 be several times continuously differentiable on the larger interval @xmath261 is often no restriction at all .
many applications of the van der corput transform have @xmath18 and @xmath28 be polynomials , exponentials , logarithms , or other @xmath262 functions that exist on large domains .
+ the bounds on the derivatives of @xmath123 and @xmath1 in condition @xmath231 part ( iv ) allow us to use estimates on stationary phase integrals that will be discussed in section [ sec : lemstationaryphase ] . by expecting the properties of @xmath123 and @xmath1 to extend to the larger interval @xmath261
, we can also extend the integrals under consideration to this larger interval .
+ ( while typical stationary phase results require only that @xmath1 is twice continuously differentiable , we require three times in order for the functions @xmath263 and @xmath264 to exist . )
3 . to understand the complex system of inequalities in condition @xmath231 part ( iv )
, it is helpful to rewrite them in terms of taylor approximations .
the first three inequalities then become @xmath265 for @xmath250 with implicit constant @xmath17 in the second and third lines .
the equality on line gives definite form to our earlier statement that the larger @xmath211 can be , the more linear @xmath123 appears at the point @xmath81 .
+ alternately , one can interpret @xmath211 as somehow representing the rate of decay as one takes successive derivatives . at the point @xmath266
, the various inequalities imply among other things that @xmath267 and @xmath268 . + the conspicuous absense of a constant @xmath269 is intentional .
we could include such a constant , but later estimates are made slightly simpler by having @xmath270 here are some examples of the function @xmath211 in different circumstances .
in each case we assume that the associated constants always take the same values : @xmath239 equals @xmath12 , and @xmath233 , @xmath234 , @xmath235 , @xmath236 , @xmath237 , and @xmath238 all equal @xmath271 . *
if @xmath18 and @xmath28 are polynomials , there exists an @xmath272 such that condition @xmath231 is satisfied with @xmath273 and @xmath274 for sufficiently large @xmath81 . *
if @xmath18 and @xmath28 are exponential functions ( @xmath275 , with @xmath226 ) , there exists an @xmath272 such that condition @xmath231 is satisfied with @xmath276 and @xmath274 . *
if @xmath277 and @xmath278 , which corresponds to the riemann zeta function at @xmath279 , then there exists an @xmath280 such that condition @xmath231 is satisfied with @xmath281 and @xmath274 independent of @xmath282 . *
if @xmath283 and @xmath4 , then there exists an @xmath280 such that condition @xmath231 is satisfied with @xmath284 and @xmath285 . *
if @xmath18 is a power function and @xmath286 for some constant @xmath16 , then there exists an @xmath272 such that condition @xmath231 is satisfied with @xmath276 and @xmath285 .
assume that condition @xmath231 holds for some function @xmath211 and @xmath232 .
let @xmath261 be as in condition @xmath231 part ( iii ) .
we assume that @xmath209 is a positive , real - valued function , bounded away from @xmath287 on the interval @xmath261 , and that @xmath28 is a real - valued function on @xmath261 as well .
( the case @xmath288 may be considered by taking the conjugate of the sum . )
this will guarantee that @xmath93 is a continuous , monotonic function , so that @xmath289 is well defined .
let @xmath290 be the union of all intervals @xmath205\subset j$ ] such that @xmath291 and @xmath292 for @xmath293 $ ] , let @xmath294 be the union of all intervals @xmath205\subset j$ ] such that @xmath295 , @xmath296 , and @xmath297 for @xmath293 $ ] , and let @xmath294 be the set of points @xmath298 such that @xmath299 but @xmath300 and @xmath301 .
it is possible that @xmath290 will contain isolated points due to @xmath302 having a @xmath287 on an interval where it is otherwise negative , and it is also possible that @xmath294 will contain isoluted points due to @xmath34 having a @xmath287 on an interval where it is otherwise non - zero ; we denote the set of isolated points in @xmath290 and @xmath294 by @xmath303 and @xmath304 respectively .
let @xmath305 and @xmath306 denote the endpoints of the non - zero - length intervals contained in their respective sets .
in particular , @xmath307 and @xmath308 . for our final assumption ,
suppose that if @xmath309 is an interval contained in @xmath290 , then either @xmath302 equals @xmath287 on the whole interval and @xmath310 does not equal @xmath287 at any point on the interval nor does it tend to @xmath287 at the endpoints , or @xmath302 does not equal @xmath287 at any point on the interval nor does it tend to @xmath287 at the endpoints ; and if @xmath309 is an interval contained in @xmath294 , then @xmath28 does not tend to @xmath287 at the endpoints of this interval .
[ thm : main ] assume all the conditions of section [ section : assumptions ] hold .
then @xmath311 where @xmath312 and @xmath313 the number @xmath314 equals the cardinality of the set @xmath315 in particular , @xmath316 also , @xmath317 equals the smallest value in the interval @xmath318 $ ] such that @xmath319 is an integer ; if no such value exists , then we may take @xmath320 .
similarly , @xmath321 equals the largest value in the interval @xmath322 $ ] such that @xmath323 is an integer , and if no such value exists , then we may take @xmath324 .
the functions @xmath211 and @xmath232 are as in condition @xmath231 , the sets @xmath290 , @xmath303 , @xmath294 , and @xmath304 are all as in section [ section : assumptions ] , and the functions @xmath325 , @xmath326 , @xmath327 , @xmath328 , @xmath329 , @xmath330 , @xmath331 , @xmath332 , and @xmath333 are all as in section [ section : notation ] .
the implicit constants in the big - o terms are dependent upon the constants @xmath233 , @xmath234 , @xmath235 , @xmath236 , @xmath237 , @xmath238 , @xmath239 in condition @xmath231 .
the convergence of the integrals in @xmath223 will be guaranteed by proposition [ prop : integrability ] before continuing to further refinements of the main theorem , let us pause to assure the reader that the vast collection of error terms above are more psychologically daunting than they are computationally difficult . as an instructive example , consider the case of corollary [ cor : iwakowimprovement ] , where @xmath334 @xmath335 , and @xmath336 .
@xmath73 and @xmath337 are assumed to be positive and @xmath16 and @xmath69 both strictly greater than @xmath17 . without loss of generality , we may assume @xmath338 ( otherwise , we would instead consider the right - hand side of ) . note that @xmath18 and @xmath28 are both power functions , so condition @xmath231 holds with @xmath233 , @xmath234 , @xmath235 , @xmath236 , @xmath237 , and @xmath238 all equal to @xmath271 , @xmath239 equal to @xmath12 , @xmath211 equal to @xmath339 for some small @xmath272 dependent on @xmath16 only , and @xmath232 equal to @xmath28 .
the benefit of our conditions @xmath340 is that on the interval @xmath341 $ ] , the functions @xmath209 and @xmath211 are both bounded below by some constant ( depending on @xmath16 and the implicit constant in @xmath342 ) .
this immediately gives that the @xmath343 , @xmath344 , and @xmath345 terms are all @xmath346 , since each summand of these three terms contains a factor of @xmath232 divided by some positive power of @xmath209 and @xmath211 , which are bounded from below .
( note that @xmath347 terms only appear if @xmath95 is bigger than @xmath209 to begin with . ) for the @xmath222 terms , we need to consider @xmath317 and @xmath321
. it might be quite difficult to calculate these values explicitly , but we only make the @xmath222 terms bigger by assuming @xmath317 is as small as possible ( that is , @xmath348 ) and @xmath321 is as large as possible . since @xmath349 on @xmath341
$ ] , we have that @xmath350 and @xmath351 .
thus , by first applying the fact that @xmath352 and @xmath349 and then that @xmath350 , the integral in @xmath353 may be bounded as @xmath354 which is @xmath355 again .
a similar argument shows that the remaining terms in @xmath353 and all the terms in @xmath356 are @xmath355 as well .
this is one reason why the error terms are much easier to deal with than at first glance : _ provided @xmath352 and @xmath357 on @xmath3 $ ] , the error terms @xmath343 and @xmath358 for @xmath359 are all bounded by @xmath360}u(x))$ ] , where the implicit constant is dependent on the implicit constants in the lower bounds of @xmath123 and @xmath24 .
_ to nicely bound the first integral in @xmath223 , however , we must use the fact that @xmath211 and @xmath209 are not only bounded below , but also somewhat large .
in particular , we have @xmath361 the greatest complication comes in evaluating the remaining condition from section [ section : assumptions ] dealing with the functions @xmath362 and @xmath363 and the remaining error terms in @xmath223 . in this case , the function @xmath364 is never zero on the interval @xmath365 $ ]
. hence @xmath294 and @xmath366 are empty , so both @xmath367 and the final sum of @xmath223 are @xmath287 .
this also shows that the functions @xmath2 and @xmath1 satisfy the remaining condition from section [ section : assumptions ] . since @xmath2 and @xmath1 are power functions , we can quickly verify , with simple hand calculations , that the functions @xmath363 , @xmath362 , @xmath368 , and @xmath369 are all power functions as well ; that is , each is of the form @xmath370 .
in particular , the functions @xmath371 are equal to @xmath372 where @xmath373 is a constant depending on @xmath16 and the sign of @xmath374 . the functions @xmath375 are equal to @xmath376 with new constants dependent on @xmath16 and the sign of @xmath374 . also ,
@xmath302 is a power function ; so , @xmath290 is either the interval @xmath377 $ ] or is empty ( if @xmath378 for all @xmath81 , in which case the remaining @xmath379 terms equal @xmath287 and we are finished ) . in general , we expect that if @xmath28 and @xmath209 share the same rate of decay in their derivatives , then the various functions @xmath363 , @xmath362 , and the @xmath380 and @xmath172 functions should all be relatively well - behaved . in our case , since @xmath1 and @xmath123 are both power functions , taking derivatives of each entails multiplying by a constant and dividing by @xmath81 . thus , we have that @xmath381 is a constant multiple of @xmath382 , hence why @xmath363 is a power function as well .
returning our attention to the remaining error terms of @xmath223 , first consider the integrals .
we have @xmath383 the functions @xmath384 and @xmath385 have at most @xmath386 zeroes on @xmath3 $ ] , and since the functions @xmath387 and @xmath388 are decreasing and since @xmath303 is empty , the remaining terms in @xmath223 are bounded by @xmath389 ( we implicitly used that @xmath70 and @xmath71 implies @xmath390 . ) since we assumed @xmath338 , we have @xmath391 , this completes the proof of corollary [ cor : iwakowimprovement ] . under fairly strong and yet quite common conditions
, we can remove much of the complication from the assumptions of section [ section : assumptions ] and from @xmath223 .
[ thm : alternate4 ] assume that @xmath2 and @xmath1 be real - valued functions satisfying @xmath392 on @xmath3 $ ] and condition @xmath231 for a function @xmath211 such that @xmath393 on @xmath3 $ ] .
also assume that @xmath394 , where , as in theorem [ thm : main ] , @xmath314 is the cardinality of the set @xmath315 under these conditions , the result of theorem [ thm : main ] holds , with @xmath223 now just equal to @xmath395 the condition that @xmath394 also simplifies more of the terms to the following . @xmath396 using theorem [ thm : alternate4 ] , we may now prove corollary [ thm : karatsubakorolevimprovement ] . for corollary [ thm : karatsubakorolevimprovement ] recall that the conditions are that @xmath18 and @xmath28 are real - valued functions , @xmath18 four times continuously differentiable and @xmath28 two times continuously differentiable on @xmath3 $ ] . also , @xmath397 for @xmath398 on @xmath3 $ ] , for constants @xmath75 , @xmath399 , and @xmath35 .
first , we wish to remove from consideration the case when @xmath400 for @xmath401 equal to @xmath149 or @xmath43 .
suppose that not only is satisfied for both @xmath149 and @xmath43 , but that @xmath402that is , the nearest integer to @xmath188 and @xmath44 are the same .
in this case , we have by the mean value theorem that @xmath403 for some @xmath404 $ ] .
but by our assumption on the size of @xmath405 and @xmath406 , we have that @xmath407 .
combining these we have that @xmath408 .
thus we have @xmath409 we can add in the sum @xmath410 to the right - hand side , since this sum by assumption has at most one term , which is of size @xmath411 .
thus the corollary is proved in this case , since @xmath412 ad @xmath413 are bounded by @xmath411 .
if @xmath414 does not equal @xmath415 , then we may replace an endpoint @xmath401 where holds with @xmath416 the closest value to @xmath401 inside the interval @xmath3 $ ] where @xmath417 or @xmath418 ; by the same argument as the previous paragraph , doing this generates an error of size @xmath419 .
the only way we could not do this is if @xmath420 and @xmath209 is at least @xmath421 at some point @xmath81 on @xmath3 $ ] , in which case , @xmath422 , so that @xmath423 and the corollary follows in this case by the same argument as in the last paragraph .
thus it suffices now to prove corollary [ thm : karatsubakorolevimprovement ] in the case where @xmath424 or @xmath90 for @xmath401 equal to @xmath149 and @xmath43 .
the conditions of theorem [ thm : alternate4 ] hold with @xmath425 on @xmath3 $ ] .
the size of the constants in condition @xmath231 occur in the bounds on the derivatives of @xmath2 and @xmath1 in terms of the constants @xmath25 , @xmath24 , and @xmath35 , and @xmath232 just equals @xmath35 .
now , we go through the error terms of theorems [ thm : main ] and [ thm : alternate4 ] .
first , we have that @xmath426 for the @xmath345 terms , we use that @xmath427 and @xmath51 to obtain @xmath428 for the @xmath429 terms , we again need to understand @xmath317 and @xmath321 . in the proof of corollary [ cor : iwakowimprovement ] , we used the fact that @xmath123 and @xmath24 were generally quite large to get good bounds . here instead , when @xmath123 is small , we shall show that @xmath430 and @xmath431 are going to be very large most of the time .
in particular , by the mean value theorem , for some @xmath404 $ ] we have @xmath432 since @xmath319 must be an integer .
thus for @xmath353 , we have the following bound .
@xmath433 a similar bound holds for @xmath356 .
now for the last term @xmath223 , we have the following much simpler calculation .
@xmath434 this completes the proof of corollary [ thm : karatsubakorolevimprovement ] .
we conclude with two additional theorems which extend the results of theorem [ thm : main ] in specific directions .
the following theorem shows that if the various error terms from theorem [ thm : main ] are nicely bounded as @xmath43 tends to infinity , then many of them can be replaced by a constant plus a @xmath435 error .
[ thm : toinfinity ] assume all the conditions of section [ section : assumptions ] hold with @xmath149 fixed and @xmath43 tending to infinity .
let @xmath436 be the set of all @xmath37 $ ] such that @xmath437 , where @xmath211 is as in condition @xmath231 .
let @xmath438 for a set @xmath439 have the usual meaning of the boundary of the set ( the endpoints of all intervals and the isolated points ) .
suppose we have the following functions @xmath440 } \frac{u(x)}{f''(x)(b - x)^3}\left(1+\frac{1}{f''(x)m(x)}+\frac{1}{m(x)(b - x)}\right ) \ dx \\ & \qquad\qquad + \sum_{x\in \partial ( k_b\cap[a,\overline{b } ] ) } \frac{u(x)}{f''(x)^2(b - x)^3}+\delta_2(b)\\ & \qquad \int_{k_b } \frac{u(x)}{f''(x ) m(x)^3}\left ( 1+\sqrt{f''(x)}m(x ) \right ) \left(1 + \frac{1+|m'(x)|}{f''(x ) m(x ) } \right)\ dx\\ & \qquad \qquad + \sum_{x\in \partial k_b } \frac{u(x)}{f''(x)^2m(x)^3}\left(1+\sqrt{f''(x)}m(x)\right),\end{aligned}\ ] ] and that @xmath441 , @xmath442 , and @xmath443 all tend to @xmath287 for some sequence of @xmath43 s tending to @xmath150 ( the sequence may be different for different functions ) .
the function @xmath345 and the value of @xmath321 are as defined in theorem [ thm : main ] .
in addition , assume that the sums and integrals in @xmath353 , @xmath444 , and the @xmath379 terms all converge as @xmath43 tends to infinity in theorem [ thm : main ] .
then @xmath445 as @xmath149 remains fixed and @xmath43 tends to infinity , where @xmath116 is some constant and @xmath446 and all other functions are as in theorem [ thm : main ]
. the sets @xmath447 , @xmath448 , and @xmath449 all refer to the similarly named sets from section [ section : assumptions ] corresponding to the set @xmath450 instead of the set @xmath3 $ ] .
the implicit constants in the big - o terms are dependent upon the constants @xmath233 , @xmath234 , @xmath235 , @xmath236 , @xmath237 , @xmath238 , @xmath239 in condition @xmath231 .
an analogous result holds if @xmath43 is fixed and @xmath149 tends to @xmath451 .
much of the accuracy gleaned from theorems [ thm : main ] and [ thm : alternate4 ] comes when @xmath209 is small .
if @xmath209 is large , instead , then @xmath343 shifts from being an explicit term to a big - o term .
however , with additional work it can be made somewhat explicit .
[ thm : refinement ] assume the conditions of section [ section : assumptions ] hold .
suppose that @xmath452 and let @xmath453 and @xmath454 be real numbers satisfying @xmath455 also let @xmath456 and @xmath457
. then we may replace @xmath458 in the statement of theorem [ thm : main ] or [ thm : alternate4 ] with @xmath459 where @xmath460 can be estimated multiple different ways . * if @xmath461 , then @xmath462 * if @xmath463 , then @xmath464 * if @xmath463 and @xmath465 , then @xmath466 suppose that @xmath463 and @xmath467 .
we can then optimize the above results , so that @xmath468 equals @xmath469 in all cases , the implicit constants depend on the constants associated to condition @xmath231 and on the implicit constants in the bounds on @xmath453 and @xmath454 .
this theorem holds true if @xmath149 is everywhere replaced by @xmath43 .
the results of theorems [ thm : toinfinity ] and [ thm : refinement ] can be combined to provide a more explicit @xmath470 term as @xmath43 tends to infinity .
it is crucial to the van der corput transform to have very good evaluations of integrals that contain stationary phase points .
the most powerful results known to the author are those of redouaby and sargos in @xcite .
[ lem : redsar1](lemma 2 in @xcite ) consider a function @xmath471\to \mathbb{r}$ ] that is @xmath472 , and postive real numbers @xmath473 @xmath474 @xmath475 @xmath476 @xmath477 @xmath478 @xmath479 such that @xmath480 \end{array}\ ] ] moreover , let @xmath481 be positive , and define @xmath482 if @xmath483 , then @xmath484 where @xmath485 if @xmath486 , then @xmath487 the implicit constant is dependent upon the @xmath453 s and @xmath239 .
the restriction @xmath241 in condition @xmath231 guarantees that @xmath488 .
see lemma 4 in @xcite . also , the above result which may be seen as a one - sided form of huxley s result in @xcite is a corollary to a more general result ( lemma 6 in @xcite ) , which makes the term @xmath489 explicit .
redouaby and sargos include the additional assumption that @xmath490 are both @xmath491 ; however , this condition is never used in the lemmas we cite , and so may be safely ignored .
[ lem : redsar2](lemma 9 in @xcite ) in addition to the hypothesis of the previous lemma , suppose that we have a function @xmath492\to \mathbb{c}$ ] that is @xmath493 , and suppose there exist positive real numbers @xmath494 such that @xmath495 then @xmath496 this result still holds if we weaken the assumptions of the previous lemma so that @xmath2 is only @xmath493 instead of @xmath472 , and the condition on @xmath33 may likewise be ignored .
[ prop : main1 ] suppose the hypotheses of lemmas [ lem : redsar1 ] and [ lem : redsar2 ] hold and suppose @xmath497 . then @xmath498 and @xmath499 where @xmath500 we just apply lemma [ lem : redsar2 ] followed by lemma [ lem : redsar1 ] .
the advantage of our condition @xmath231 is now the following .
[ prop : main2 ] suppose condition @xmath231 holds , that @xmath2 has a stationary phase point at @xmath298 , and that @xmath501\subset [ x - m(x),x+m(x)]\cap j$ ] .
then the conditions of lemmas [ lem : redsar1 ] and [ lem : redsar2 ] hold with @xmath502 and with @xmath234 , @xmath233 , @xmath235 , @xmath236 , @xmath237 , @xmath238 and @xmath239 equal to the similarly named constants in condition @xmath231 .
morever , with these definitions , the error terms become @xmath503 where if @xmath504 ( resp . , @xmath505 ) then the first ( resp . , second ) big - o term disappears .
however , we will also need good estimates on oscillatory integrals without a stationary phase point .
the first and second derivative tests below are well - known and often accompany any discussion of the van der corput transform . following them we give some more exact estimates that make use of the integration by parts technique we described in section [ section : heuristics ] .
[ lem:1stderivtest ] ( first derivative test lemma 5.1.2 in @xcite ) let @xmath18 be real and differentiable on the open interval @xmath506 with @xmath93 monotone and @xmath507 on @xmath506 .
let @xmath28 be real , and let @xmath29 be the total variation of @xmath28 on the closed interval @xmath508 $ ] plus the maximum modulus of @xmath28 on @xmath508 $ ] . then @xmath509 [ lem:2ndderivtest ] ( second derivative test
lemma 5.1.3 in @xcite ) let @xmath18 be real and twice differentiable on the open interval @xmath506 with @xmath510 on @xmath506 .
let @xmath28 be real , and let @xmath29 be the total variation of @xmath28 on the closed interval @xmath508 $ ] plus the maximum modulus of @xmath28 on @xmath508 $ ] .
then @xmath511 [ prop:2ndderivtest ] ( condition @xmath231 version of first and second derivative tests ) suppose condition @xmath231 holds .
if for some @xmath512 , we have @xmath508\subset i_x:=j \cap [ x - m(x),x+m(x)]$ ] , then @xmath513 } |f'(x)|}\ ] ] and @xmath514 with uniform implicit constant .
we can bound @xmath29 in lemmas [ lem:1stderivtest ] and [ lem:2ndderivtest ] by @xmath515 using condition @xmath231 . also by condition @xmath231
, we have that the minimum of @xmath516 on @xmath248 is at worst @xmath517 . in some cases , the first derivative test needs to be made explicit ,
for which we have the following result . [
prop : intbyparts ] let @xmath518)$ ] , and @xmath519)$ ] , and define @xmath520 by @xmath521 suppose that @xmath522 on an interval @xmath508 $ ] , and let @xmath523 where the sum ranges over all @xmath524 $ ] where @xmath525 .
then we have @xmath526_\alpha^\beta \\ & \qquad + o(k_r(\alpha,\beta ) ) + o ( |h_r(\alpha)|)+o ( |h_r(\beta)|)\end{aligned}\ ] ] we begin by applying integration by parts twice to our original integral . @xmath527_\alpha^\beta \\ & \qquad-\int_\alpha^\beta \frac{1}{2\pi i } \frac{(f'(x)-r)g'(x)-g(x)f''(x)}{(f'(x)-r)^2 } e(f(x)-rx ) \ dx \notag\\ & \qquad= \left[\frac{g(x)}{2\pi
i ( f'(x)-r)}e(f(x)-rx)\right]_\alpha^\beta \notag \\ & \qquad\qquad -\left [ \frac{1}{(2\pi i)^2}h_r(x ) e(f(x)-rx)\right]_\alpha^\beta \label{eq : propintbyparts1}\\ & \qquad\qquad + o\left ( \int_\alpha^\beta \left| \frac{d}{dx } h_r(x ) \right| \ dx \right ) \label{eq : propintbyparts2}\end{aligned}\ ] ] the two terms on line are bounded trivially by @xmath528 .
the remaining integral on line is bounded by the total variation of @xmath520 , which is at most the modulus of @xmath520 at every critical point added to @xmath529 and @xmath530 .
@xmath531 is , by definition , the sum of the moduli @xmath520 at every critical point ( since the derivative by the assumption that @xmath522 on @xmath508$]always exists ) , so the proof is complete . by making use of the following proposition
, we can sometimes avoid the @xmath532 term in proposition [ prop : intbyparts ] .
an alternate form of the resulting error terms can be found in lemma 5.5.5 of @xcite .
[ prop : fbound ] suppose condition @xmath231 holds .
if @xmath533 , then for all @xmath81 such that @xmath534 and all @xmath172 such that @xmath535 we have @xmath536 by taylor s remainder theorem , we can write @xmath537 } f^{(3)}(y ) \right ) \label{eq : f'bound}\end{aligned}\ ] ] with implicit constant 1 . by condition @xmath231
, we have @xmath538 } |f^{(3)}(y)| \le \frac{\eta f''(x)}{m(x)}.\ ] ] therefore the big - o term in is at most @xmath539 . since @xmath243 and
since @xmath540 is either @xmath287 or shares the same sign as @xmath197 , this gives the result .
[ prop : intbypartsvar ] suppose @xmath2 and @xmath1 satisfy condition @xmath231 for some function @xmath211 and its associated constants .
let @xmath541 .
assume that @xmath10 is not in the interval from @xmath81 to @xmath542 , and that @xmath543 is farther from @xmath10 than @xmath81 is .
then we have that @xmath544_{y = x+\epsilon \cdot m(x)}^x \\ & \qquad + o\left ( \frac{u(x ) f''(x)}{|f'(x)-r|^3 } \right)+ o\left ( \frac{u(x)}{m(x)(f'(x)-r)^2}\right)\end{aligned}\ ] ] we also have the following bound : @xmath545_{y = x+\epsilon \cdot m(x)}=o\left ( \frac{u(x)}{m(x)f''(x ) } \right).\ ] ] as in proposition [ prop : intbyparts ] , we apply integration by parts twice to our starting integral @xmath546_{y = x+\epsilon\cdot m(x)}^x\notag\\ & \qquad\qquad + o(h_r(x))+o(h_r(x+\epsilon\cdot m(x)))\notag \\ & \qquad\qquad + o\left ( \int_{x+\epsilon\cdot m(x)}^x \left|\frac{d}{dy } \frac{(f'(y)-r)g'(y)-g(y)f''(y)}{(2\pi i)^2(f'(y)-r)^3 } \right| \ dy \right ) \label{eq : intbypartsvar}\end{aligned}\ ] ] the relation @xmath547_{y = x+\epsilon \cdot m(x ) } = o\left ( \frac{g(x)}{m(x)f''(x ) } \right)\ ] ] holds by proposition [ prop : fbound ] .
using condition @xmath231 , we have the following additional bounds .
@xmath548 the last equality holds since @xmath549 is bigger than @xmath550 by our assumption about the relative placement of @xmath81 , @xmath10 , and @xmath542 : for example , if @xmath551 , then @xmath552 . finally , to estimate the integral in first notice that @xmath553 on the interval between @xmath81 and @xmath543 , the maximum of @xmath554 is @xmath555 and since we are integrating over an interval of length @xmath211 , the contribution of this term is at most @xmath556 similarly , one can show that @xmath557 we estimate the remaining integral by using @xmath558 : @xmath559_{x+\epsilon \cdot m(x)}^x \right ) \\ & = o\left(\frac{g(x ) f''(x)}{|f'(x)-r|^3}\right ) .
\qedhere\end{aligned}\ ] ] if @xmath28 is a constant , then we may remove the term @xmath560 from the statement of proposition [ prop : intbypartsvar ] .
this would remove the term @xmath561 from @xmath345 in theorem [ thm : main ] .
however , very often one finds a term of this same size occuring in @xmath429 regardless .
the primary technique we shall use to evaluate sums will be a variant the euler - maclaurin summation formula ; however , the euler - maclaurin formula itself sums over all integers in an interval @xmath3 $ ] , while we will often want to sum over all values @xmath81 in some interval @xmath562 $ ] for which @xmath563 is an integer ( for some function @xmath564 ) .
[ lem : emsum ] ( euler - maclaurin summation , first derivative version page 10 in @xcite ) suppose @xmath2 is a differentiable function on @xmath508 $ ] then @xmath565 [ prop : inversesum ] suppose @xmath362 is a real - valued , differentiable function on @xmath3 $ ] .
suppose @xmath564 is a differentiable function such that @xmath566)=[\alpha,\beta]$ ] , then @xmath567 } \left|s(x)\cdot g(x ) \right| \right)\end{aligned}\ ] ] where @xmath568 is a sum over all @xmath569 for @xmath570 .
we apply the first derivative version of euler - maclaurin summation to each interval where @xmath564 is monotonic ( and hence , @xmath571 is 1 - 1 ) .
then we apply a change of variables with @xmath572 .
the sum @xmath573 } \left|s(x)\cdot g(x ) \right| \right)\ ] ] bounds the contributions of @xmath574 and @xmath575 arising from the endpoints of the intervals of monotonicity .
[ lem : poissonsummation ] ( lemma 5.4.2 in @xcite and @xmath43 are integers , so we provide the proof when @xmath149 , @xmath43 are real.]poisson summation ) let @xmath576)$ ] for real numbers @xmath577 , then @xmath578 we rewrite the euler - maclaurin summation formula as @xmath579 but we also have @xmath580 by integration by parts , which completes the proof .
we end this section with two more specific propositions .
the first will give us good bounds on the partial sums and tails of @xmath163 , our modified sawtooth function , as well as guarantee its convergence .
the second will show that the integrals in theorem [ thm : main ] involving the @xmath380 and @xmath172 functions will converge .
[ prop : fouriertail ] suppose @xmath581 , @xmath582 $ ] , and @xmath583 .
then @xmath584 where the implicit constant is uniform .
also , the sum is convergent as @xmath196 tends to @xmath150 and the same bounds hold in this case . without loss of generality
, we may replace that @xmath81 with @xmath585 and so may assume that @xmath586 .
we start with the assumption that @xmath196 is finite .
next , we remove all appearances of the @xmath197 from the sum .
we use @xmath587 and @xmath588 to denote constants which only depend on the variables @xmath16 , @xmath81 , and @xmath197 .
@xmath589 here , @xmath587 contains the terms where @xmath590 or @xmath591 .
in particular , it is no larger than @xmath592 .
similarly , @xmath588 is dominated by the sum of terms @xmath593 , and hence is also @xmath594 .
now we pause a moment to show that we may let @xmath595 with no problems of convergence .
the sum @xmath596 is , up to a constant multiplier , the fourier series for the sawtooth function @xmath597 and converges for all @xmath81 .
this implies that @xmath598 converges for all @xmath16 and @xmath81 .
therefore , since @xmath599 we have that @xmath600 converges for all @xmath16 , @xmath81 , and @xmath197 . if @xmath601 then @xmath602 and hence @xmath603 so we may assume for the remainder of the proof that @xmath604 .
now we remove the absolute value in the condition on the sum and apply summation by parts .
@xmath605 since we assumed @xmath606 $ ] , we have that @xmath607 . for the numerator
, we use that @xmath608 .
this gives the following estimate , which completes the proof .
[ prop : integrability ] suppose the conditions of section [ section : assumptions ] hold , then the function @xmath610 is integrable on @xmath294 and the function @xmath611 is integrable on @xmath290 .
recall that @xmath171 the functions @xmath363 and @xmath362 are continuous and bounded on @xmath261 .
the functions @xmath327 and @xmath328 are continuous inside @xmath294 , and the functions @xmath329 and @xmath330 are continuous inside @xmath290 .
the only possible barrier to integrability is the presence of a zero in the denominator of the function at an endpoint of an interval .
the only terms in the denominator of @xmath612 are @xmath1 and @xmath123 , which are both bounded away from @xmath287 on @xmath294 .
the only terms in the denominator of @xmath327 are @xmath1 and @xmath123 again , but @xmath613 has a factor of @xmath614 in the denominator ; however , in @xmath615 , the @xmath614 in the denominator of @xmath613 is cancelled by the factor of @xmath614 in the numerator of @xmath616 , so this too has no zeroes in the denominator at the endpoints of @xmath294 .
the case when we are considering an interval @xmath309 where @xmath617 but @xmath310 does not equal nor tend to @xmath287 is analogous .
so we will assume we are in the case where @xmath618 does not equal nor tend to @xmath287 on the interval in question for the remainder of this proof .
we have @xmath619 and @xmath620 in addition , @xmath621
so , the only terms in the denominator of @xmath264 are @xmath364 and @xmath622 , and the only terms in the denominator of @xmath263 are @xmath623 and @xmath622 as all the @xmath364 terms cancel
. the technique required will change slightly depending on whether we consider the @xmath624 terms or the @xmath625 terms . as @xmath362 tends to @xmath287 ( which is equivalent to @xmath364 tending to @xmath287 )
, @xmath626 will tend to either @xmath627 or @xmath287 . by the assumptions of section [ section : assumptions ] , if @xmath362 tends to @xmath287 at the endpoint of an interval in @xmath290 , then @xmath363 can not tend to @xmath287 at the same point .
if the sign is chosen so @xmath623 tends to @xmath627 , then the terms @xmath623 and @xmath622 in the denominator of @xmath628 do not vanish , and the two copies of @xmath364 in the denominator of @xmath629 are cancelled by the two copies of @xmath364 in the numerator of @xmath371 . therefore , in this case , there are no zeroes in the denominator . for the remaining case , when @xmath630 , we have that @xmath631 where the implcit constant depends on the size of @xmath363 near this point . in particular ,
since @xmath1 and @xmath123 do not approach @xmath287 at the endpoints of @xmath290 , equation implies that a copy of @xmath623 tending to @xmath287 in the denominator of a function can be cancelled by a copy of @xmath364 in the numerator to prevent the presence of a zero in the denominator .
in particular , @xmath371 will not have a zero in the denominator . by
, the only term in @xmath632 that could produce a zero in the denominator is the derivative @xmath633 therefore , it suffices to show that has no zeroes in the denominator at an endpoint of @xmath290 . by applying to line and noting that in this case @xmath634 , we have @xmath635 since @xmath622 does not tend to @xmath287 on @xmath290 , the derivative has no zeroes in the denominator .
we begin by applying poisson summation ( lemma [ lem : poissonsummation ] ) .
this gives @xmath636 we wish to alter the end - points of some of these integrals .
define @xmath637 , the new left endpoint , by @xmath638 and @xmath639 , the new right endpoint , by @xmath640 note that @xmath641 is positive , if and only if @xmath10 lies to the left of @xmath149 and vice - versa .
the transformation from @xmath3 $ ] to @xmath642 $ ] has the effect that if @xmath10 is close to but not equal to an endpoint @xmath149 or @xmath43 , then we shift that endpoint away from @xmath10 by a distance @xmath643 or @xmath644 , respectively . in the statement of theorem [ thm : main ] , @xmath314 is defined by @xmath645 the values @xmath646 and @xmath647 thus count the number of @xmath172 such that @xmath648 and @xmath649 , respectively . after altering the end - points , we have @xmath650 since there are only finitely many values of @xmath172 for which @xmath648 and @xmath651 , the sum on line is finite and so we may omit the limit and sum from @xmath451 to @xmath150 .
first , consider those integrals in the sum on line arising from @xmath652 .
if @xmath653 or if @xmath654 , there are no such integrals , and so the total contribution of this ( null ) set of integrals is @xmath287 . otherwise , there are at most @xmath271 such integrals , each of which by proposition [ prop:2ndderivtest ] ( our variant of the first and second derivative tests ) is bounded by @xmath655 thus the contribution of these integrals is @xmath656 .
there can only be additional integrals in line if @xmath657 , so we will assume as such when bounding them .
consider next those integrals from @xmath637 to @xmath149 arising from @xmath658 .
there are @xmath659 such integrals , each of which , by proposition [ prop:2ndderivtest ] is bounded by @xmath660 so the total contribution of these integrals is bounded by @xmath661
lastly , we consider the sum of those integrals in line arising from @xmath662 . in particular ,
let @xmath663 denote the sum in question .
we apply proposition [ prop : intbypartsvar ] to each integral in @xmath664 to obtain @xmath665 here , each sum is over all @xmath172 satisfying @xmath662 . for the sum on line , we apply a change of variables @xmath666 , where @xmath667 , and then apply proposition [ prop : fouriertail ] .
the sum is then bounded by @xmath668 , since we are under the temporary assumption that @xmath657 .
the sum on line has at most @xmath669 terms , so is bounded by @xmath670 by euler - maclaurin summation ( lemma [ lem : emsum ] ) , the sum on line is bounded by @xmath671 thus the total contribution of the integrals from @xmath637 to @xmath149 in line arising from @xmath672 is bounded by @xmath673 where @xmath674 similarly the contribution of the integrals from @xmath639 to @xmath43 in line is bounded by @xmath675 therefore , @xmath676 suppose that @xmath677 .
we will need to estimate the integral at this particular @xmath172 in separately .
we shall break the integral into a piece surrounding the stationary phase point and an integral away from the stationary phase point .
in particular , we write @xmath678 where @xmath679 . since @xmath680 is to the left of @xmath43 , we have that @xmath639 either equals @xmath43 or @xmath681 . by propositions [ prop : main1 ] and [ prop :
main2 ] , we have @xmath682 here we used the bound @xmath683 similarly , if @xmath684 , then we dissect the integral from @xmath637 to @xmath43 in in a similar way , obtaining , @xmath685 where @xmath686 . for all remaining integrals with a stationary phase point in
, we may assume that @xmath687 .
as before , we denote the stationary phase point corresponding to a given @xmath172 by @xmath10 . for each such @xmath688
, we write the corresponding integral in line as @xmath689 where @xmath690 and @xmath691 . by construction , @xmath692=i_{x_r}\cap [ a_r , b_r]$ ] with @xmath248 as in the statement of condition @xmath231 . for this section , we shall focus on the contribution of the middle terms .
consider @xmath693 applying propositions [ prop : main1 ] and [ prop : main2 ] with @xmath694 to @xmath695 , this becomes @xmath696 the sum in line added to the first terms in lines and if they exist sum to @xmath697 the main term in the van der corput transform . to begin evaluating the sum in line , consider values @xmath172 such that @xmath698 . by the mean value theorem , we can write @xmath699 for some @xmath700 . by the bounds of condition @xmath231 ,
the @xmath172 under consideration must also satisfy @xmath701 , which , in turn implies that @xmath702 .
therefore , we split the sum over @xmath703 in line into two pieces .
the first piece , over all @xmath172 such that @xmath704 , has at most @xmath646 terms , with each term bounded by @xmath705 thus the total contribution of these terms is at most @xmath706 since @xmath707 . the second piece , over all remaining @xmath172 between @xmath188 and @xmath44 , is bounded by @xmath708 where @xmath317 is the smallest value in the interval @xmath318 $ ] such that @xmath319 is an integer .
note that if no such integer exists , then this sum is empty .
we now apply proposition [ prop : inversesum ] to with @xmath709 and @xmath710 on the interval @xmath711 $ ] . using the bounds from condition @xmath231 liberally
, we see that is bounded by @xmath712 since @xmath8 on @xmath3 $ ] , @xmath713 has no local extrema on this interval , so the final term from proposition [ prop : inversesum ] does not appear .
we have a similar bound for the second sum in line : @xmath714 we apply proposition [ prop : inversesum ] again to the sum on line to get @xmath715 now consider @xmath716 note that the remaining terms of @xmath695 in line as well as the second terms from lines and if they exist appear in @xmath717 .
we have thus far shown that @xmath718 for each integral in @xmath717 , we apply proposition [ prop : intbyparts ] .
this gives @xmath719 where in each sum it is assumed that @xmath720 .
we start estimating the first sum of and write @xmath721 there are at most @xmath646 terms on lines and , each with size @xmath722 by proposition [ prop : fbound ] , for a total contribution of @xmath723 we can evaluate the sum on line explicitly in some cases .
suppose @xmath724 .
if @xmath725 , then we pull out the term where @xmath726 , so that the remaining terms on line form a modified sawtooth function .
in particular , the sum on line becomes @xmath727 which equals @xmath468 , when @xmath724 .
otherwise , if @xmath728 , we may bound the sum on line using proposition [ prop : fouriertail ] by @xmath729 . recall that @xmath468 is not explicit when @xmath728 , and instead is just a big - o approximation with one of its terms of the form @xmath730 . by a similar argument , one can show that all the sums on line are equal to @xmath731 we break the sum from into three smaller sums like so @xmath732 the sum on the right - hand side of line is @xmath733 if @xmath734 or @xmath735 , then the smallest @xmath736 could be in this sum is @xmath737 ; otherwise , we will have a term where @xmath738 .
thus this sum is bounded by @xmath739 by using condition @xmath231 and proposition [ prop : fbound ] , each term on lines and has size at most @xmath740 and there are at most @xmath741 such terms for a total contribution of at most @xmath742 by a similar argument both the sums on line are bounded by @xmath743 .
for the first sum on line , we have @xmath744 equals @xmath637 or @xmath745 .
the sum of terms with @xmath746 are bounded by the sum in line , which we just showed has size @xmath747 , so we only need to estimate the terms where @xmath748 . by proposition [ prop : fbound ] and
the bounds of condition @xmath231 , each such term is bounded by @xmath749 we may therefore bound the sum over those @xmath172 with @xmath748 by a sum of terms @xmath750 for all @xmath751 $ ] and obtain @xmath752 } \frac{g(x_r)}{f''(x_r)^2 m(x_r)^3 } = o(\delta_4)\ ] ] by the same argument as in line . therefore
, we have shown that @xmath753 the proof will therefore be finished when we show that the sum of all the @xmath532 terms is bounded by @xmath754 .
the @xmath532 s are the sum of terms of the size @xmath755 at points @xmath81 where the derivative with respect to @xmath81 vanishes .
we may safely ignore points where the derivative does not exist , since all the intervals that give rise to the @xmath532 terms do not contain stationary phase points .
the derivative of @xmath520 is ( using @xmath756 as shorthand for @xmath540 ) @xmath757 we set the numerator equal to 0 and solve for @xmath756 .
first , suppose @xmath299 .
if @xmath34 also equals @xmath287 , then the numerator is @xmath287 if and only if @xmath758 also equals @xmath287 ( @xmath209 is never @xmath287 , and we assumed @xmath759 at all points in consideration ) ; but if @xmath28 and @xmath758 equal @xmath287 , then @xmath520 also equals @xmath287 , so these points contribute nothing to @xmath532 .
thus the only contribution from points @xmath81 where @xmath299 come when @xmath760 and @xmath300 , i.e. , those points in @xmath366 as defined in section [ section : assumptions ] . at these points , solving for @xmath756 yields @xmath761 , so the contribution to the @xmath532 terms from these points is bounded by .
@xmath762 for the remainder of this section , we will suppose then that @xmath763 .
there is no way to make the numerator equal @xmath287 if @xmath764 , since both @xmath28 and @xmath209 are assumed to be non - zero , or if @xmath765 , since all the above functions are real - valued ; otherwise , the solution is given by @xmath766 where @xmath290 and @xmath294 are as defined in section [ section : assumptions ] .
plugging this value of @xmath756 into @xmath767 , we obtain @xmath768 where @xmath769
if we ignore the constraint that @xmath172 be an integer for the moment , then we can imagine that determines functions @xmath770 the contribution of the @xmath532 terms is then at most @xmath771 we now apply proposition [ prop : inversesum ] to see that the contribution of the @xmath532 terms is bounded by @xmath772 this completes the proof of theorem [ thm : main ] .
we follow the proof of theorem [ thm : main ] , until section [ section : remainingintegrals ] . as in the conditions of theorem [ thm : alternate4 ] , suppose that @xmath393 for all @xmath773 $ ] and that @xmath394 . then we may use proposition [ prop : intbypartsvar ] in place of proposition [ prop : intbyparts ] to the integrals in @xmath717 and obtain @xmath774 we used @xmath149 and @xmath43 in place of @xmath637 and @xmath639 due to @xmath646 and @xmath647 being zero .
likewise , we have no @xmath744 , @xmath775 terms since by our assumption on @xmath211 , @xmath776 and @xmath777 for all @xmath172 .
just as in the proof of theorem [ thm : main ] , the terms in line equal @xmath778 the terms on lines and we bound as we did the terms in line , obtaining @xmath779 since we have no @xmath532 terms , we can let @xmath223 just equal @xmath780 and this completes the proof of theorem [ thm : alternate4 ] in this case .
the proof of this theorem mostly entails following the proof of theorem [ thm : main ] and simply being more careful with how certain error terms arise . in section [ section : prooffirststep ] , we want to show that we can replace @xmath781 with some constant @xmath782 . but
these terms simply serve as a bound for the sum @xmath783 which , for fixed @xmath643 , is a constant as @xmath43 tends to @xmath150 .
( if @xmath43 is near @xmath149 , we may need to take @xmath643 to be smaller , thus it s not always a constant . )
likewise , the term @xmath784 that appear in section [ section : proofhalf ] arise from bounding the term @xmath785 from proposition [ prop : main1 ] applied with the @xmath116 and @xmath43 of the statement of the proposition equal to @xmath149 and @xmath775 , respectively .
provided @xmath43 is large enough , @xmath775 will be constant , so the term @xmath785 may be likewise replaced by a constant @xmath786 .
the error terms from section [ section : prooffull ] arise from bounding the sum @xmath787 where @xmath788 ( respectively , @xmath789 ) is the error from proposition [ prop : main1 ] applied with @xmath149 ( respectively , @xmath43 ) and @xmath116 in the statement of the proposition equal to @xmath744 ( respectively , @xmath775 ) and @xmath10 .
we would like to replace the @xmath43 in line with @xmath150 .
consider first the @xmath788 terms .
each such term is fixed ( provided @xmath43 is sufficiently large enough that @xmath790 ) .
thus we write @xmath791 suppose now @xmath207 is chosen to be on a sequence tending to infinity such that @xmath792 tends to @xmath287 .
then we have that @xmath793 exists and is a constant @xmath794 by the assumed convergence of the integral in @xmath353 , and we have that @xmath795 by proposition [ prop : inversesum ] .
thus we have @xmath796 we must apply an additional idea to perform the same technique to the terms @xmath789 . in particular , if @xmath797 , then @xmath775 changes as @xmath43 grows .
we may replace all such terms by @xmath798 with an error of size @xmath799 . following the argument of the previous paragraph
, we have @xmath800 in section [ section : remainingintegrals ] , the first sum on line is constant for @xmath43 sufficiently large , and the second sum is just @xmath801 . by following the details of proposition [ prop : intbyparts ] that we applied to @xmath717 , we see that what we are left with at this point is @xmath802 where in the latter sum we assume we are taking the limit as @xmath803 tends to infinity of the sum with the additional restriction that @xmath804 .
let us , for ease of notation , rewrite the integrals as @xmath805\setminus i_{x_r } } , \ ] ] where as in condition @xmath231 the set @xmath806 equals @xmath807 $ ] .
we wish to replace @xmath43 by a sequence of @xmath207 s tending to infinity , such that @xmath808 tends to @xmath287 . for such
@xmath207 we have that above sum equals @xmath809\setminus i_{x_r}}\frac { h_r(x)}{2\pi i } e(f(x)-rx ) \ dx \\ & \qquad -\lim_{r\to \infty}\sum_{|r|\le r } \int_{[b_r ,
b'_r]\setminus i_{x_r}}\frac { h_r(x)}{2\pi i } e(f(x)-rx ) \ dx.\end{aligned}\ ] ] we now finish applying proposition [ prop : intbyparts ] by taking integration by parts of all the integrals and then bounding the variation . this combined with the ideas from earlier in the proof give us that the terms on and are bounded by @xmath810 this completes the proof of theorem [ thm : toinfinity ] .
hence we could replace some of the errors on lines and with @xmath811 now if the integrals on lines and converge and if both the second term in and the term on line go to @xmath287 on a sequence of @xmath43 s tending to infinity , then we can replace this sum by @xmath812 where the first series is converges ( absolutely ) to some constant , and the second is bounded by @xmath813 the reason for requesting that certain terms tend to @xmath287 on a sequence of @xmath43 s ( say @xmath814 , @xmath815 , etc . )
tending to infinity , is that we really need to replace the sum on line with @xmath816 and show that this limit is well - defined .
the first sum tends to a constant , and the second term can be bounded by the above collection of big - o terms , plus big - o of the specified terms at @xmath817 . since the latter are assumed to go to @xmath287 , the limit is well - defined . we can attempt to do the same for the right - side half - stationary phase errors ( those coming from proposition [ prop : main1 ] applied to @xmath694 and @xmath818 ) , but we run into an additional problem that some of these errors are not fixed as @xmath43 increases .
namely , those where @xmath819 .
thus let @xmath436 be the set of @xmath9 $ ] such that @xmath437 , which contains at least the point @xmath43 trivially , and let @xmath443 equal @xmath820 } \frac{u(x)}{f''(x)^2(b - x)^3}+\delta_2(b)\\ & \qquad \int_{k_b } \frac{u(x)}{f''(x ) m(x)^3}\left ( 1+\sqrt{f''(x)}m(x ) \right ) \left(1 + \frac{1+|m'(x)|}{f''(x ) m(x ) } \right)\ dx\\ & \qquad \qquad + \sum_{x\in \partial k_b } \frac{u(x)}{f''(x)^2m(x)^3}\left(1+\sqrt{f''(x)}m(x)\right).\end{aligned}\ ] ] note that @xmath821 counts the contribution of all the @xmath822 terms in line , since these terms appear only if @xmath437 .
provided @xmath443 tends to @xmath287 for some sequence of @xmath43 s tending to infinity and the first integral in @xmath223 is convergent as @xmath43 tends to infinity , we may replace the remaining contributions to lines and with a constant plus @xmath821 plus big - o of the tail of the integrals .
this removes the term @xmath823 completely , replacing it with @xmath821 .
the first integral in @xmath223 also has a contribution from the following section , but the argument here still applies under the same conditions .
we also desire to show that the @xmath379 terms can be replaced with constants plus a @xmath435 error if each of them converges as @xmath43 tends to infinity .
if @xmath824 tends to @xmath287 ( which is implied by @xmath825 tending to @xmath287 ) on a sequence of @xmath43 s tending to infinity , then indeed the @xmath379 terms can be replaced by a constant plus @xmath826 where @xmath447 and @xmath448 refer to the sets @xmath294 and @xmath290 as defined in section [ section : assumptions ] corresponding to the interval @xmath450 instead of @xmath3 $ ] . in the estimates of theorem
[ thm : main ] , we only roughly estimated the size of @xmath827 , which roughly are the first - order endpoint contributions at @xmath43 and @xmath149 , when @xmath123 was large . in this case , we included error terms of the size @xmath828 , which are at least the size of the first and last term of our initial sum .
these terms can be improved , but require additional work .
but recall , at the beginning of the proof of theorem [ thm : main ] , we bounded a sum of the type for almost all @xmath172 , the sum is in fact finite . ]
@xmath829 in line , which along with the sum in line contributed to the @xmath730 error in @xmath830 . in bounding this sum ,
we exploited the fact that the first - order endpoint contributions at @xmath149 should cancel significantly ( for example , in our estimation of line ) . we expect , however , that more should be true , that not only should the first - order endpoint contributions display cancellations , but the integrals themselves should also display cancellation near @xmath149 . as in the conditions of theorem [ thm : refinement ] , assume that @xmath831 , let @xmath453 and @xmath454 be real numbers satisfying @xmath832 and let @xmath833 and @xmath834 .
we now define a value @xmath835 similarly to how we defined @xmath637 on line . @xmath836 if @xmath837 then @xmath835 is between @xmath149 and @xmath637 , except , possibly , for two values of @xmath172 if @xmath454 is very close to @xmath214 .
we now break the original sum into several pieces : @xmath838 where the last term comes from the term @xmath839 as in . in the sequel
, we shall refer to these three groups of integrals as type i , type ii , and type iii integrals , respectively .
we will postpone fully estimating the type i integrals for now , and instead modify them into a more symmetric form .
we first replace @xmath28 by @xmath840 } g'(z ) \cdot |x - a|\right)\ ] ] and @xmath841 by @xmath842 } f^{(3)}(z)}{3 ! } |x - a|^3 \right).\end{aligned}\ ] ] doing so gives @xmath843 shifting the bounds of integration and sending @xmath641 to @xmath844 , the type i integrals can be rewritten as @xmath845 the type ii and type iii integrals we evaluate using proposition [ prop : intbypartsvar ] .
this gives us @xmath846 similar to proposition [ prop : fbound ] , we have that @xmath847 with implicit constant @xmath17 , so that @xmath848 for @xmath849 .
thus , the sum in line is bounded by @xmath850 and the sum in line is bounded by @xmath851 since @xmath852 .
the terms on line are bounded by @xmath853 finally , we apply euler - maclaurin summation to the sum on line to bound it by @xmath854 since @xmath454 is at most @xmath214 . using all of these estimates for the integrals on line ( with the negative sign now ,
as they appeared in the proof of theorem [ thm : main ] ) and adding the terms from line , we obtain the following . @xmath855 now we estimate the type i integrals and the sum on line in three separate cases . in this case , we begin by adding the @xmath172 and @xmath856 terms from the type i integrals together , obtaining @xmath857 we estimate each integral in by an application of integration by parts . @xmath858_c^0
\label{eq : case1 1 } \\ & \qquad \qquad + \int_0^c \frac{2\epsilon '
\cos(2\pi \epsilon ' x)}{f''(a)x+r } e\left ( \frac{f''(a)}{2}x^2+rx \right ) \ dx \label{eq : case1 2}\\ & \qquad \qquad - \int_0^c \frac{f''(a ) \sin(2\pi \epsilon ' x)}{\pi(f''(a)x+r)^2 } e\left ( \frac{f''(a)}{2}x^2+rx \right ) \ dx.\label{eq : case1 3}\end{aligned}\ ] ] the term @xmath859 on line is @xmath287 when @xmath601 and bounded by @xmath860 if @xmath861 .
we will then apply the first and second derivative tests ( lemmas [ lem:1stderivtest ] and [ lem:2ndderivtest ] ) to the integral on line .
the total variation may be bounded by @xmath862 since @xmath863 and @xmath864 is at least @xmath172 for @xmath865 $ ] , the maximum modulus on the interval is at most @xmath866 .
therefore the integral on line is bounded by @xmath867 we will also apply the first and second derivative tests to the integral on line . using @xmath868 , we have that @xmath869 here we used that the function @xmath870 is @xmath287 at @xmath601 , increases monotonically to a maximum at @xmath871 , and then decreases mononotonically to @xmath287 as @xmath81 tends to infinity .
thus the type i integrals are bounded by @xmath872 which gives the bound on @xmath460 in this case . also , we bound the sum in line by @xmath873 using proposition [ prop : fouriertail ] .
as in case 1 , we sum @xmath172 and @xmath856 terms and use the first derivative test , obtaining @xmath874 by the same argument , we may complete the fourier series in .
@xmath875 together these give @xmath460 in this case . in particular , by saying @xmath149 is far from an integer , we want @xmath465 .
this implicitly requires that @xmath876 . in this case
, we sum over all positive values of @xmath172
seperately from all negative values of @xmath172 . ignoring the constant multiplier @xmath877 for the moment ,
the sum over positive @xmath172 values give @xmath878 here we will apply the second derivative test to the integrals in lines . the variation plus maximum modulus in both integrals is bounded by @xmath879 so that both integrals ( as well as the corresponding integrals for the sum over negative @xmath172 ) are bounded by @xmath880 in this case , again , we bound the terms in line by @xmath881 .
but since @xmath882 , this term is dominated by @xmath883 thus we have @xmath460 in this final case .
when @xmath463 and @xmath467 , the following choices of @xmath454 and @xmath453 optimize the error terms from line : @xmath884 when @xmath885 or @xmath886 ; @xmath887 when @xmath888 ; and , @xmath889 when @xmath890 . inserting these values into the error terms for cases 2 and 3
gives the result at the end of theorem [ thm : refinement ] .
these methods work similarly at @xmath43 in place of @xmath149 .
recall that we are attempting to bound the error in the following van der corput transform .
@xmath891 we can apply the main theorems of this paper to either the sum on the left or the right of . however , we obtain best results if we use both at different times , dependent on how big @xmath892 is . in this case , we apply theorem [ thm : toinfinity ] to the left - hand side of .
thus @xmath895 @xmath896 and @xmath897 is non - zero . for @xmath233 , @xmath234 , @xmath235 , @xmath236 , @xmath237 , and @xmath238
equal to @xmath271 and @xmath239 equal to @xmath12 , we have that condition @xmath231 is satisfied for @xmath273 for some small fixed @xmath280 ( independent of @xmath73 ) and @xmath285 .
hence @xmath898 on @xmath899 $ ] , so @xmath294 is the full interval @xmath261 , and both @xmath290 and @xmath366 are empty .
we first need to check that the conditions of theorem [ thm : toinfinity ] hold . using our assumption that @xmath893
, we can estimate @xmath441 by @xmath900 thus @xmath441 tends to @xmath287 as @xmath901 and @xmath902 tends to infinity along the positive integers .
we can bound @xmath442 by @xmath903 which clearly tends to @xmath287 as @xmath73 tends to infinity .
the set @xmath436 equals @xmath904 $ ] .
let @xmath905 be the largest integer in @xmath906 $ ] for which @xmath907 is also an integer .
then the first integral and sum of @xmath443 are bounded by @xmath908 } \frac{x^{1/2}}{(n - x)^3}\left ( 1 + x^{-1/2 } + \frac{x^{1/2}}{n - x}\right ) \
dx\\ & \qquad + \frac{\overline{n}}{(n-\overline{n})^3 } + \frac{n(1+\epsilon)^{-1}}{(n - n(1+\epsilon)^{-1})^3}\\ & \ll \frac{\overline{n}^{1/2}}{(n-\overline{n})^2 } + \frac{\overline{n}}{(n-\overline{n})^3}.\end{aligned}\ ] ] however , by the mean value theorem , we have , for some @xmath909 $ ] , @xmath910 and @xmath911 , so that the first integral and sum of @xmath443 is bounded by @xmath912 likewise the final integral and sum of @xmath443 are bounded by @xmath913 thus @xmath914 is bounded by @xmath915 and converges to @xmath287 on the sequence @xmath916 , @xmath917 . finally , we need to show that the integrals and sums in @xmath353 , @xmath223 , and @xmath379 converge .
the integrand of the integral in @xmath353 is @xmath918 which converges like @xmath919 .
the integrand of the integral in @xmath223 is @xmath920 which converges like @xmath921 .
the integrand of the integral in @xmath367 is @xmath922 .
the sum in the @xmath379 term is bounded by @xmath923 , since @xmath294 is a single interval and @xmath924 never changes sign .
since @xmath925 converges like @xmath926 , this shows that the conditions of theorem [ thm : toinfinity ] are satisfied .
it also shows that @xmath927 is bounded by @xmath928 .
to apply the theorem we calculate @xmath470 explicitly by @xmath929 if @xmath894 , then @xmath930 ; and otherwise , we get that @xmath931 , since @xmath140 is greater than @xmath932 for @xmath933 , and so @xmath934 equals @xmath287 . combining the error terms together
we see that @xmath935 which completes the proof of the corollary in this case . using our assumption that @xmath893
, we can estimate @xmath441 by @xmath900 we wish to apply some of the further ideas remarked in the paper allowing us to replace some convergent integrals with constants plus a @xmath435 error .
for this we will need to understand @xmath436 , the set of @xmath37 $ ] such that @xmath437 .
but since @xmath281 , @xmath436 just equals @xmath936 $ ] .
we also wish to bound @xmath443 .
let @xmath905 be the largest integer in @xmath906 $ ] for which @xmath907 is also an integer .
then first integral and sum in @xmath443 is bounded by @xmath908 } \frac{x^{1/2}}{(n - x)^3}\left ( 1 + x^{-1/2 } + \frac{x^{1/2}}{n - x}\right ) \
dx\\ & \qquad + \frac{\overline{n}}{(n-\overline{n})^3 } + \frac{n(1+\epsilon)^{-1}}{(n - n(1+\epsilon)^{-1})^3}\\ & \ll \frac{\overline{n}^{1/2}}{(n-\overline{n})^2 } + \frac{\overline{n}}{(n-\overline{n})^3}.\end{aligned}\ ] ] however , by the mean value theorem , we have , for some @xmath909 $ ] , @xmath910 and @xmath911 , so that this first integral and sum of @xmath443 is bounded by @xmath912 likewise the final integral and sum of @xmath443 are bounded by @xmath913 thus @xmath914 is bounded by @xmath915 and converges to @xmath287 on the sequence @xmath916 , @xmath917 . to replace convergent integrals by constants
, we must also have that @xmath937 and @xmath441 tend to @xmath287 for some sequence of @xmath43 s tending to infinity .
but the terms on line are bounded by @xmath938 which tends to @xmath287 as @xmath73 tends to infinity , and as noted above there exists a sequence of @xmath73 s for which @xmath441 tends to @xmath287 . finally we need to show that the integrands themselves converge .
the integrand of the integral in @xmath353 is @xmath918 which converges like @xmath919 .
the integrand of the integral in @xmath223 is @xmath920 which converges like @xmath921 .
the integrand of the integral in @xmath367 is @xmath922 .
we can quickly see that the remaining summands in these three terms are @xmath136 as well .
the sum in the @xmath379 term is bounded by @xmath923 , since @xmath294 is a single interval and @xmath924 never changes sign .
therefore we may switch the integrals in big - o terms for a constant plus big - o of their tails . in total
, we have one explicit term , arising from @xmath470 , and a constant arising from the terms at @xmath149 and the completed integrals .
the errors we have are as follows : a term of size @xmath355 from @xmath470 when @xmath73 is @xmath112 times a perfect square ; a term of size at most @xmath939 from @xmath940 ; a term of size @xmath941 from @xmath441 and @xmath443 ; a term of size @xmath942 from the two terms on line ; a term of size @xmath939 from the term @xmath943 in @xmath944 ; and a term of size @xmath928 arising from the tail of the integrals .
thus , we have that @xmath6 equals @xmath945 where @xmath946 and @xmath116 is some fixed constant
. the terms in @xmath353 converge like @xmath947 however , we can say something more specific .
the errors which give rise to @xmath948 are unchanged as @xmath73 goes to infinity , thus this can be represented as some constant @xmath786 . likewise ,
the errors which give rise to the integral in @xmath353 are absolutely convergent as @xmath73 tends to infinity ; thus , we may let @xmath794 be the sum of all such error terms , and see that the errors which contribute to the integral in @xmath353 equal @xmath949 thus we may replace the term @xmath950 with @xmath951 for some constant @xmath952 .
this type of argument is generally referred to as a tail - swapping argument .
let @xmath905 be the largest integer in @xmath906 $ ] for which @xmath907 is also an integer .
we can bound @xmath356 by @xmath953 however , by the mean value theorem , we have , for some @xmath909 $ ] , @xmath954 and @xmath911 , so that @xmath955 finally , we have that @xmath956 by a similar tail - swapping argument , we can replace the term @xmath754 with @xmath957 so in total , @xmath6 equals @xmath945 where @xmath946 and @xmath116 is some fixed constant .
now we apply theorem [ thm : refinement ] to @xmath959 times the conjugate of the right - hand side of .
we use the conjugate to guarantee that @xmath209 is positive .
we have @xmath960 @xmath961 since @xmath362 is never @xmath287 on @xmath3 $ ] and @xmath962 is always positive on @xmath3 $ ] we have that @xmath294 is empty and @xmath963 $ ] .
so , to complete our useful definitions , we have @xmath964 and again @xmath629 never equals @xmath287 on @xmath3 $ ] .
again , for @xmath233 , @xmath234 , @xmath235 , @xmath236 , @xmath237 , and @xmath238 equal to @xmath271 and @xmath239 equal to @xmath12 , we have that condition @xmath231 is satisfied for @xmath273 for some small fixed @xmath280 independent of @xmath73 and @xmath965
. since @xmath149 is fixed , we have @xmath966 by theorem [ thm : refinement ] , we have that @xmath470 equals @xmath967 by our assumption that @xmath73 is an integer , @xmath968 and so @xmath969 for @xmath441 note that in this case @xmath44 is always an integer , and hence @xmath970 equals @xmath287 .
therefore , we have @xmath971 for @xmath972 we may assume @xmath973 and @xmath974 ( at worst , this assumption only makes the @xmath222 terms larger ) .
thus , @xmath975 and @xmath976 to estimate @xmath223 , note that by our above definitions we have @xmath977 and so , @xmath978 thus we have shown that , in this case @xmath979 by undoing our conjugation , multiplying by @xmath141 , and moving this term to the correct side of the equation .
in all the main results and the various examples given in this paper , the results tend to be at their best when @xmath980 ( for @xmath401 equal to @xmath149 or @xmath43 ) is large compared with @xmath981 .
however , when @xmath980 is on the order of @xmath981 , the estimates tend to be dominated by an inexplicit term , and even when @xmath980 is much smaller than @xmath981 , we still do not have great asymptotics , such as in corollary [ thm : exampleerror ] .
could the error term @xmath6 in this range be made more explicit ?
if one s notion of explicit is a closed form that can be estimated moderately fast on a computer , then yes ; if one s notion of explicit requires good asymptotic data , then the answer so far appears to be no .
suppose that @xmath209 is generally smaller than @xmath17 on @xmath3 $ ] .
the largest contributions to @xmath6 come from bounding an integral of the form @xmath982 where @xmath983 .
we will assume for a moment that @xmath10 is in the interval @xmath3 $ ] . in the proof of theorem [ thm : main ] , we bounded this integral with two techniques : first , possibly shifting @xmath401 to @xmath984 if @xmath985 was smaller than @xmath986 , and second , applying propositions [ prop : main1 ] and [ prop : main2 ] , our variants of redouaby and sargos results .
however , another result of redouaby and sargos provides an alternate bound for such an integral . under the conditions of lemma [ lem : redsar1 ] and the additional assumption that @xmath987 to simplify the error terms , we have the following result ( combining lemma 3 and lemma 9 in @xcite ) for @xmath988 : @xmath989 where @xmath990 , @xmath991 and @xmath992 denotes the modified fresnel integral function or @xmath993 where we have used @xmath994 and would further break the function up into its real and imaginary parts .
] , @xmath995 an analogous result holds for the integral from @xmath116 to @xmath43 .
let us assume once again that @xmath996 .
just as in propositions [ prop : main1 ] and [ prop : main2 ] , we see an error term of the form @xmath997 .
this new estimate on stationary phase integrals is useful in the sense that the remaining error terms are of order of magnitude smaller than the explicit term provided @xmath71 ,
@xmath25 is of order much larger than @xmath17 , and @xmath998 is of order much smaller than @xmath999 this in turn is useful to the van der corput transform if the good range for @xmath998 includes @xmath1000 , and this will be true provided @xmath1001 is of an order of magnitude much greater than @xmath1002 . to apply these new estimates ,
we need to make a few further small changes in the proof of theorem [ thm : main ] .
first , for this @xmath172 , we do not change @xmath401 to @xmath1003 , removing the term @xmath1004 from @xmath1005 .
second , since we only apply the above estimates to an integral on one side of @xmath10 , we obtain terms of size @xmath1006 from the terms now left uncancelled when we apply propositions [ prop : main1 ] and [ prop : main2 ] to the other side . finally , since we do not have any integral from @xmath401 to @xmath1003 in @xmath717 , all the associated terms no longer need to be bounded .
if @xmath1007 , then this means we lack the term @xmath1008 as well as @xmath1009 and @xmath1010 .
this thus removes the term with @xmath1011 from @xmath1012 , and reduces the last pair of terms in @xmath1013 to @xmath1014 when @xmath10 is not in the interval @xmath3 $ ] , say @xmath1015 , then we need to be slightly more careful . we will need to assume that condition @xmath231 holds on the larger interval @xmath1016 $ ] , and then consider the difference of integrals @xmath1017 we would apply propositions [ prop : main1 ] and [ prop : main2 ] to the first integral and the above estimates to the second
while this may seem like a great improvement over earlier theorems and in a sense , it is it also is very hard to apply .
the region @xmath1018 requires us to understand the function @xmath992 in the region @xmath1019 .
unfortunately , while there are very good asymptotics for @xmath992 when @xmath1020 tends to @xmath287 or infinity , there are no good asymptotics when @xmath1021 . if @xmath1022 $ ] .
( if @xmath10 is not in @xmath3 $ ] , then we technically need to bound an integral that runs from @xmath1023 or @xmath745 whichever is inside the integral @xmath3 $ ] to @xmath401 .
the argument works the same either way but in this case we need to be careful that all our conditions and assumptions still apply outside the interval , otherwise the term @xmath1024 is meaningless . )
what is the behavior of the error term for @xmath1025 in the region @xmath131 ?
the most significant contribution to @xmath6 in this case comes from an integral like @xmath1026 for @xmath1027 close to @xmath140 .
this integral could be made more explicit in the following way : under the conditions of lemma [ lem : redsar1 ] , we have the following result ( lemma 3 in @xcite ) for @xmath988 : @xmath1028 where @xmath990 and @xmath1029 when @xmath1030which in our case corresponds to @xmath131the terms @xmath1031 and @xmath1032 are of the same order of magnitude
. the first term on the right hand side of may also be rewritten in terms of the fresnel integral functions @xmath1033 and @xmath1034 .
this can make the integral in line , and hence the majority of @xmath6 , explicit in the sense that we would have a term in closed form with much smaller big - o errors ; however , this relies on knowing values of the fresnel integral functions when @xmath81 is on the order of @xmath17 , and in this region , we have no good asymptotics for the fresnel integral functions and even direct computation is difficult .
we may rewrite such an integral in terms of the fresnel integral functions @xmath453 and @xmath439 in the following way : @xmath1035
in section [ sec : variation ] , we do not make use of the fact that the intervals taken as arguments in the @xmath532 not only fail to include the point @xmath10 but also are excluded from an interval around @xmath10 .
is there a simple way to take this into account and does doing so improve the results in a substantial way ? can the results of this paper be explained in terms of the geometry of curves , in the same way as @xcite ?
dekking and mends - france conjecture in @xcite that @xmath1036 , the curve related to the exponential sum @xmath1037 is resolvable .
can the results of this paper be used to prove this ?
the author acknowledges support from national science foundation grant dms 08 - 38434 `` emsw21-mctp : research experience for graduate students , '' under which the investigations into the van der corput transform began .
a. fedotov and f. klopp .
renormalization of exponential sums and matrix cocycles . in _ sminaire : quations aux drives partielles .
20042005 _ , pages exp .
cole polytech . ,
palaiseau , 2005 . r. r. moore and a. j. van der poorten . on the thermodynamics of curves and other curlicues . in _
miniconference on geometry and physics ( canberra , 1989 ) _ , volume 22 of _ proc .
centre math .
_ , pages 82109 .
univ . , canberra , 1989 . | we improve the error term in the van der corput transform for exponential sums @xmath0 for many functions @xmath1 and @xmath2 , we can extract the next term in the asymptotic , showing that previous results , such as those of karatsuba and korolev , are sharp . of particular note ,
the methods of this paper avoid the use of the truncated poisson formula , and thus can be applied to much longer intervals @xmath3 $ ] with far better results .
we provide a detailed analysis of the error term in the case @xmath4 and @xmath5 . |
multiferroic materials , in which ferromagnetic , ferroelectric , and/or ferroelastic orderings coexist , have attracted significant research interest since many years @xcite due to their potential applications in data storage media , spintronics and ferroelectric random - access memories .
the co - existence of ferro-orders in multiferroics allows the possibility that the magnetization can be tuned by an applied electric field and vice versa . among the limited choices offered by the multiferroic materials , bifeo@xmath4 ( bfo ) , one of the single - phase multiferroic materials at room temperature , exhibits the co - existence of ferroelectric ordering with curie temperature ( t@xmath8 ) of 1123 k and antiferromagnetic ( afm ) ordering with a nel temperature ( t@xmath9 of 643 k @xcite . in bifeo@xmath4 ,
magnetic ordering is of antiferromagnetic type , having a spiral modulated spin structure ( smss ) with an incommensurate long - wavelength period of 62 nm @xcite .
this spiral spin structure cancels the macroscopic magnetization and inhibits the observation of the linear magnetoelectric effect @xcite .
these problems ultimately limit the use of bulk bifeo@xmath4 in functional applications .
recent investigations demonstrated that in order to perturb the smss and improve the magnetic properties of bifeo@xmath4 , partial substitution of bi by rare - earth ions @xcite or alkaline - earth ions @xcite and also substitution of fe by transition metal ions @xcite is an effective route .
it is also reported that simultaneous minor substitution of bi and fe in bifeo@xmath4 by ions such as la and mn , la and ti , nd and sc etc .
, respectively @xcite also enhanced the magnetism and ferroelectricity in bifeo@xmath4 .
recently , we have observed that simultaneous substitution of gd and ti in place of bi and fe , respectively in bifeo@xmath4 multiferroics improved their morphological , dielectric and magnetic properties at room temperature @xcite .
later on , another group also observed fascinating magnetic , optical and dielectric properties in this gd and ti co - doped bifeo@xmath4 ceramic system at room temperature @xcite .
therefore , in this investigation , we were interested to conduct experiments on temperature dependence of magnetic properties of gd and ti co - doped bi@xmath0gd@xmath1fe@xmath2ti@xmath3o@xmath4 ( x = 0.00 - 0.20 ) multiferroic materials .
in particular , our interest is to investigate the effect of temperature on remanent magnetization , coercive fields and exchange bias ( eb ) fields of these bulk polycrystalline materials .
notably , the eb phenomenon which manifest itself by a shift of a magnetic hysteresis loop in systems containing ferromagnetic / antiferromagnetic bilayers has been studied extensively since many years due to its importance in spintronic applications @xcite .
the eb effect usually occurs when the system is cooled down in an external magnetic field through the magnetic ordering temperatures .
most of the research in this field is focused on specially prepared systems , however , recent investigations reported the presence of eb effect in perovskite manganites @xcite , cobaltites @xcite , heusler alloys @xcite and multiferroics @xcite .
although eb has been observed in various bulk materials , the effect in most cases has been limited to far below room temperature ( @xmath10100 k)@xcite , thus making the systems less attractive for applications . in our previous investigation
, the magnetization vs. magnetic field ( @xmath11 ) hysteresis loops indicated the presence of exchange bias effect in the gd and ti co - doped bi@xmath0gd@xmath1fe@xmath2ti@xmath3o@xmath4 ( x = 0.00 - 0.20 ) ceramics at room temperature although the biasing field was very small @xcite .
therefore , in the present investigation , we have explored exchange bias effect in this material system at different temperatures ranging from 20 k to 300 k. we have also investigated the influence of cooling magnetic fields at temperatures 150 k and 250 k to observe a tunable exchange bias in sample composition bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 ( this composition is referred as sample x = 0.10 throughout the manuscript ) .
the polycrystalline samples having compositions bi@xmath0gd@xmath1fe@xmath2ti@xmath3o@xmath4 ( x = 0.00 - 0.20 ) were synthesized by using standard solid state reaction technique as was described in details in our previous investigation @xcite .
the powder materials and ceramic pellets sintered at 1100 k were used for magnetic characterization .
the @xmath11 hysteresis loops of bi@xmath0gd@xmath1fe@xmath2ti@xmath3o@xmath4 ( x = 0.00 - 0.20 ) multiferroic ceramics were carried out at different temperatures using a superconducting quantum interference device ( squid ) magnetometer ( quantum design mpms - xl7 , usa ) .
the temperature dependent magnetization measurements were carried out both at zero field cooling ( zfc ) and field cooling ( fc ) processes @xcite .
the magnetic hysteresis loops were also carried out at different temperatures by cooling down the sample from 300 k in various cooling fields ( @xmath6 ) @xcite .
hysteresis loops of sample x = 0.10 ( composition bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 ) carried out at different temperatures .
( b - h ) an enlarged view of the low - field m - h hysteresis loops of sample x = 0.10 obtained at temperatures : ( b ) 20 k ( c ) 50 k ( d ) 100 k ( e ) 150 k ( f ) 200 k ( g ) 250 k and ( h ) 300 k.,width=302 ]
for magnetic characterization , the m - h hysteresis loops of bi@xmath0gd@xmath1fe@xmath2ti@xmath3o@xmath4 ( x = 0.00 - 0.20 ) samples were measured at different temperatures ranging from 20 k to 300 k with an applied magnetic field of up to @xmath1250 koe . as a typical example ,
the m - h loops of sample x = 0.10 ( composition bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 ) at different temperatures were presented in figure [ fig1 ] ( a ) . figures [ fig1 ] ( b - h ) demonstrate an enlarged view of the low - field m - h hysteresis loops of this sample measured at temperatures : ( b ) 20 k ( c ) 50 k ( d ) 100 k ( e ) 150 k ( f ) 200 k ( g ) 250 k and ( h ) 300 k. the hysteresis loops for the two other compositions were shown in the supplemental figures 1(s1 ) and 2(s2 ) @xcite .
it should be noted that the m - h hysteresis loops ( figure [ fig1 ] , supplemental figures 1(s1 ) and 2(s2 ) ) were carried out initially without applying any cooling magnetic field .
the coercive fields ( h@xmath13 ) and remanent magnetization ( m@xmath14 ) extracted from the hysteresis loops were quantified as : @xmath15 , where h@xmath16 and h@xmath17 are the left and right coercive fields @xcite and m@xmath18 = @xmath19(m@xmath20-m@xmath21)@xmath19/2 where m@xmath20 and m@xmath21 are the magnetization with positive and negative points of intersection with h = 0 , respectively @xcite .
calculated values of h@xmath13 and m@xmath18 are plotted as a function of temperature in figures [ fig2 ] ( a and b ) respectively for bi@xmath0gd@xmath1fe@xmath2ti@xmath3o@xmath4 ( x = 0.00 - 0.20 ) samples .
both the coercive fields and remanent magnetizations are higher for sample x = 0.10 than those for x = 0.00 and x = 0.20 . therefore , it is clear that the substitution of 10% ti in place of fe in bi@xmath0gd@xmath1feo@xmath4 compound significantly increased h@xmath13 and m@xmath18 .
however , a further increment of ti to 20% in place of fe reduced h@xmath13 and m@xmath18 although their net values are still higher than that of ti undoped bi@xmath0gd@xmath1feo@xmath4 sample .
as was reported in our previous investigation @xcite , the larger values of h@xmath13 and m@xmath14 in sample x = 0.10 are related with the microstructure of the composition i.e. with homogeneous small grain size of the material . in the case of la and nb co -
substituted bifeo@xmath4 @xcite and pr and zr co - substituted bifeo@xmath4 compounds @xcite , increase in coercive field with the substitution was also attributed to decrease in grain size .
the value of h@xmath13 in these materials are much larger than that in pure bifeo@xmath4 @xcite and such a large value of coercivity in gd and ti co - substituted bifeo@xmath4 samples may be related to their magnetic anisotropy @xcite . ) and ( b ) remanent magnetization ( m@xmath14 ) as a function of temperature in bi@xmath0gd@xmath1fe@xmath2ti@xmath3o@xmath4 ( x = 0.00 - 0.20 ) compounds . both h@xmath13 and
m@xmath14 are higher over a wide range of temperatures in sample x = 0.10.,width=340 ] the remarkable feature observed from figure [ fig2](a ) is that the coercivity of these ceramic samples increases with temperature .
figure [ fig7 ] shows , as a typical example for x = 0.10 sample , that while the coercivity increases with increasing temperature , the maximum magnetization ( m@xmath22 ) at 50 koe decreases with increasing temperature @xcite . for all studied samples , the coercivity show a strong temperature dependency and the h@xmath13 values are significantly higher at room temperature than that at 20 k. for example , the h@xmath13 value at room temperature for sample x = 0.10 is three times higher than that at 20 k. the usual trend in a typical magnetic system is for the coercivity to increase with decreasing temperature @xcite since the anisotropy decreases much more sharply than does the magnetization with increasing temperature @xcite . in similar multiferroic material systems , e.g. la doped bifeo@xmath4 @xcite and lufe@xmath23o@xmath24 @xcite , the h@xmath13 values were found to increase at low temperature than that at room temperature .
the unexpected decreasing trend of h@xmath13 at low temperature as compared with the room temperature values for all compositions investigated here may be explained by the presence of magnetoelectric coupling in these multiferroic materials @xcite .
the presence of the magnetoelectric coupling produces additional contribution to the anisotropy that actually acts to decrease the effective magnetic anisotropy @xcite .
if @xmath25 is the uniaxial anisotropy constant in the absence of coupling effects and @xmath26 is the uniaxial anisotropy constant in the presence of magnetoelectric coupling , then @xmath27 + + here @xmath28 is the homogeneous magnetoelectric co - efficient that is related to the dzyaloshinsky moriya magnetic field , @xmath29 is the magnetic susceptibility in the direction perpendicular to the antiferromagnetic vector , @xmath30 is the spontaneous electric polarization @xcite .
thus the temperature variation in h@xmath13 is determined by the competition between the magnetic anisotropy and the magnetoelectric coupling @xcite .
( the maximum magnetization at 50 koe ) and h@xmath13 as a function of temperature in sample x = 0.10 ( bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4).,width=340 ] the higher values of h@xmath13 and m@xmath14 in sample x = 0.10 than those of x = 0.00 and 0.20 compositions motivated us to investigate further its magnetic properties in details .
therefore , we have carried out the temperature dependence of the magnetization ( @xmath31 ) of bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 bulk sample .
the @xmath31 curves measured in zfc and fc modes in the presence of 500 oe applied magnetic field is shown in figure [ fig3 ] . in zfc process
, the sample is initially cooled from 300 k to the lowest achievable temperature and data were collected while heating in the presence of the applied field . on the other hand , in the fc mode , data values are collected while cooling in the presence of the magnetic field which is commonly known as ` cooling magnetic field ' @xcite . here
the temperature dependence magnetization measurements demonstrate clearly that both zfc and fc curves of bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 ceramic coincide with each other without showing any bifurcation which indicates the absence of any spin flipping effect @xcite . in similar multiferroic materials both the zfc and fc curves were also found to coincide with each other @xcite .
curves ) of bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 sample measured in zfc and fc processes in the presence of 500 oe applied magnetic fields .
both zfc and fc curves coincide with each other without showing any bifurcation.,width=340 ] + as shown in figures 1(a - h ) , the room temperature @xmath11 hysteresis loop as well as the loops taken at other temperatures exhibit an asymmetric shift towards the magnetic field axes @xcite .
this is a signature of the presence of an exchange bias effect in multiferroic bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 material @xcite . in the present investigation , the hysteresis loops of gd and ti co - doped bifeo@xmath4 ceramic system is unsaturated even with an applied magnetic field of up to @xmath1250 koe . which confirm the basic antiferromagnetic nature of the compounds .
notably , undoped bifeo@xmath4 possesses a very narrow hysteresis loop with a very small but non - zero remanent magnetization ( 0.0009 emu / g ) and a coercive field of @xmath32110 oe at room temperature @xcite . compared to pure bifeo@xmath4 , the center of m - h loops of bi@xmath0gd@xmath1fe@xmath2ti@xmath3o@xmath4 ( x = 0.00 - 0.20 ) compounds are wider ( as shown typically in the enlarged view of sample x = 0.10 , figures [ fig1 ] ( b - h ) ) which suggests a weak ferromagnetic nature @xcite of this co - doped ceramic system .
the weak ferromagnetic nature of this material system is also revealed from temperature dependent magnetization curves @xcite . in this way
we anticipate the co - existence of strong - anisotropic ferri / ferromagnetic ( fm ) and antiferromagnetic domains in this multiferroic material system . as a consequence of the exchange coupling at interfaces between these multiple magnetic domains
, it is expected that the system acts as a natural system for generating eb effect in bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 multiferroics @xcite . ) as a function of temperature in bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 .
the h@xmath33 values were calculated from the asymmetric shift of the @xmath34-@xmath35 hysteresis loops of figures [ fig1](b - h ) .
the hysteresis loops were taken without applying a cooling magnetic field.,width=9 ] the exchange bias field ( h@xmath33 ) from the loop asymmetry along the field axis can be quantified as @xmath36 where @xmath37 and @xmath38 are the left and right coercive fields , respectively @xcite .
the variation of h@xmath33 as a function of temperature in bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 calculated from the asymmetric shift of the @xmath34-@xmath35 hysteresis loops of figures [ fig1](b - h ) is shown in figure [ fig4 ] .
the temperature dependence of h@xmath33 for two other compositions were presented in the supplementary information ( supplementary table 1 ) @xcite . as was mentioned earlier , the eb effect usually occurs when the system is cooled down in an external magnetic field through the nel temperature ( t@xmath39 ) .
it is worth mentioning that in bifeo@xmath4 ceramic system the exchange bias effect was observed without any magnetic - field - annealing process through t@xmath39 @xcite which is the conventional method of inducing unidirectional anisotropy @xcite .
it has also been observed without using any alloy layers @xcite , however , then the biasing strength of bifeo@xmath4 is observed to be very weak with @xmath7 = 36 oe @xcite at room temperature .
the exchange bias fields indicate the strength of the exchange coupling of an exchange bias system .
as shown in figure [ fig4 ] , values of @xmath7 of bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 at 150 k and 250 k are higher than that at any other temperatures .
the @xmath7 values shown in figure [ fig4 ] is observed without applying any cooling magnetic field and therefore the biasing strength is weak and random . in the case of a fm / afm bilayer system , it is well known that cooling magnetic field plays an important role in establishing a strong unidirectional anisotropy due to the exchange coupling @xcite .
hence , in the following experiments , the magnetic hysteresis loops of bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 material were carried out at temperatures 150 k and 250 k by cooling down the sample from 300 k in various cooling magnetic fields ( @xmath6 ) ranging from 20 koe to 60 koe . in each experiment related to cooling magnetic fields ,
the measuring magnetic fields were from -30 koe to 30 koe .
the details of the loop asymmetry at different cooling fields and temperatures can be found in the supplementary figures 3(s3 ) and 4(s4 ) @xcite . ) at 150 k and 250 k for bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 ceramic .
the graph also demonstrates the dependency of @xmath7 on temperature .
, width=340 ] the influence of cooling magnetic fields on exchange bias effect at temperatures 150 k and 250 k in bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 multiferroic material is illustrated in figure [ fig5 ] .
the @xmath7 values increased significantly upon the application of cooling magnetic fields .
for example , at temperature 250 k and cooling magnetic field 60 koe , the biasing field increased more than twelve times than that for a cooling magnetic field of 20 koe . during
the field cooling experiments , the @xmath7 values reduced with increasing temperature , figure [ fig5 ] , which is similar to results reported in previous investigations @xcite .
the @xmath7 values are higher than those in related material systems @xcite specifically at high temperatures . in the present investigation ,
the temperature dependent zfc and fc measurements ( figure [ fig3 ] ) demonstrate the absence of any bifurcation from 300 k down to 5 k and therefore , we had conducted the field cooling experiments by cooling the sample from 300 k down to 150 k and 250 k. by using the squid magnetometer ( quantum design mpms - xl7 , usa ) , it was not possible to heat the sample above 300 k ; therefore , the field cooling experiments were not possible to conduct by cooling down this sample through t@xmath39 which is around 520@xmath40 k for bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 ceramic measured from temperature dependent dielectric measurements ( data not shown here ) . as was mentioned earlier , the eb has been observed in various bulk materials , however , the effect in most cases has been limited to low temperatures @xcite .
the observation of eb up to room temperature is of interest from the perspective of practical applications .
therefore , in future we intend to conduct field cooling experiments at room temperature by cooling bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 material through t@xmath39 which is above the room temperature .
the influence of @xmath6 on coercive fields were also measured at temperatures 150 k and 250 k and the results are displayed in table 1 .
notably , in figures [ fig2 ] and [ fig7 ] , we have shown the effect of temperature on @xmath5 without applying any cooling magnetic field . here in table 1
, we have inserted the effect of cooling magnetic field on @xmath5 and the measurements were carried out at two different temperatures close to room temperature .
the exchange bias fields are strongly affected by cooling magnetic field as shown in figure [ fig5 ] , however , the @xmath5 values ( table 1 ) are influenced merely by a small extent .
this is not unusual , in the case of nife / co bilayers , similar effect was previously observed , i.e. the exchange bias field is strongly affected by cooling magnetic field whereas the @xmath5 is weakly influenced @xcite .
.the table shows the effect of cooling magnetic fields on @xmath5 of bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 sample at temperatures 150 k and 250 k. notably , the effect of temperatures on @xmath5 of bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 sample without applying cooling magnetic fields was inserted in figure [ fig2 ] .
[ tab1 ] [ cols= " < ,
< , < , < , < " , ] in our previous investigation @xcite , we have synthesized nanoparticles of the same composition i.e. of bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 with a mean size of 40 - 100 nm directly from their bulk powder by using the sonication technique described in ref .
@xcite . in this specially prepared nanoparticle system , we have observed exchange bias effect and the magnitude of the exchange bias fields were also found to increase with cooling magnetic fields @xcite . obviously , the cooling field dependence of @xmath7 values of bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 bulk system were weaker than those of specially prepared bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 nanoparticles .
for example , in both bulk and nanoparticle system the highest @xmath7 values were observed at temperature 150 k by applying 60 koe cooling magnetic field . at this temperature and cooling magnetic field ,
the @xmath7 value of the bulk system is 20 % less than that of nanoparticles having sizes 40 - 100 nm .
this is worth noting as the preparation of bulk materials is comparatively easy and straightforward considering the multistep processing for the preparation of nanoparticles .
we have observed a strong influence of temperature on coercive fields and exchange bias fields of gd and ti co - doped bifeo@xmath4 multiferroics .
the coercive fields of this multiferroic system were enhanced anomalously with increasing temperature .
the anomalous enhancement of the coercivity with increasing temperature and particularly the observation of the high values of @xmath5 near / at room temperature may be of future use in potential applications where a coercivity stability at high temperature is crucially effective .
we have also observed the presence of exchange bias effect due to the interface exchange coupling between fm and afm domains of gd and ti co - doped bifeo@xmath4 material system .
the magnitude of the exchange bias fields in bulk bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 were also found to increase with cooling magnetic field .
this magnetically tunable exchange bias obtained in bulk multiferroic bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 material up to temperatures 250 k is promising , as most of the bulk materials show eb only far below room temperature .
we anticipate that the presence of both exchange and magnetoelectric coupling in bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 multiferroic material might be worthwhile for potential applications in novel multifunctional devices .
this work was supported by the world academy of sciences ( twas ) , ref.:14 - 066 rg / phys / as - i ; unesco fr : 324028567 and jsps kakenhi ( grant no .
26810117 ) .
a part of this work was conducted in the institute for molecular science ( ims ) , supported by nanotechnology platform program ( molecule and material synthesis ) of the ministry of education , culture , sports , science and technology ( mext ) , japan .
the authors thank to mr .
motoyasu fujiwara at the institute of molecular science ( ims ) for his assistance during squid measurement .
spaldin n a and fiebig m 2005 _ science _ * 309 * 5733 eerenstein w , mathur n d and scott j f 2006 _ nature _ * 442 * 759 cheong s w and mostovoy m 2007 _ nature materials _ * 6 * 13 ramesh r 2009 _ nature _ * 461 * 1218 wu s m , cybart s a , yi d , parker j m , ramesh r and dynes r c 2013 _ phys .
lett . _ * 110 * 067202 fischer p , polomska m , sosnowska i and szymanski m 1980 _ j. phys .
_ c : _ solid state phys .
_ * 13 * 1931 ederer c and spaldin n a 2005 _ phys .
_ b * 71 * , 060401(r)-4 .
yuan g l , wing s , liu j m and liu z g 2006 _ appl .
lett . _ * 89 * 052905 khomchenko v a , paixao j a , costa b f o , karpinsky d v , kholkin a l , troyanchuk i o , shvartsman v v , borisov p and kleeman w 2011 _ crystal research and technology _ * 46 * 238 khomchenko v a and kiselev d a 2007 _ appl .
lett . _ * 90 * 242901 xu q y , zai h f , wu d , qiu t and xu m x 2009 _ appl .
_ * 95 * 112510 chang f g , zhang n , yang f , wang s x and song g l 2007 _ j. phy . _
d : _ appl .
_ * 40 * 7799 anjum g , kumar r , mollah s , shukla d k , kumar s and lee c g 2010 _ j. app .
phys . _ * 107 * 103916 agarwal r a , ashima s s and ahlawat n 2012 _ j. phys .
_ d : _ appl .
phys . _ * 45 * 165001 rao t d and saket asthana 2014 _ j. appl
. phys . _ * 116 * 164102 basith m a , kurni o , alam m s , sinha b l and ahammad b 2014 _ j. appl
. phys . _ * 115 * 024102 kumar m , sati p c and chhoker s 2014 _ j mater sci : mater electron _ * 25 * 5366 dieny b , speriosu v s , parkin s s p , gurney b a , wilhoit d r and mauri d 1991 _ phys . rev . _
b * 43 * 1297 zhang t and dressel m 2009 _ phys .
_ b * 80 * 014435 pradheesh r , nair h s , sankaranarayanan v and sethupathi k 2012 _ app . phys .
* 101 * 142401 nayak a k , nicklas m , chadov s , khuntia p , shekhar c , kalache a , baenitz m , skourski y , guduru v k , puri a , zeitler u and felser j m d c c 2015 _ nature materials _ * 14 * 679 tian z m , yuan s l , zheng x f , jia l c , huo s x , duan h n and liu l 2010 _ app
_ * 96 * 142516 yu p , lee j s , okamoto s , rossell m d , huijben m , yang c h , he q , zhang j x , yang s y , lee m j , ramasse q m , erni r , chu y h , arena d a , kao c c , martin l w and ramesh r 2010 _ phys .
lett . _ * 105 * 027201 ang r , sun y p , luo x , hao c y , zhu x b and song w h 2008 _ j. app . phys . _
_ 104 _ 023914 basith m a , khan f a , ahammad b , kubota s , hirose f , -ngo d t , tran q -h and mlhave k 2015 * j. app . phys . *
* 118 * 023901 in supplemental figure 1(s1 ) ( a ) we show @xmath11 hysteresis loops of bi@xmath0gd@xmath1fe@xmath2ti@xmath3o@xmath4 ( x = 0.00 ) sample measured at different temperatures .
figure 1(s1 ) ( b - h ) demonstrate the enlarged view of the low field hysteresis loops at temperatures : ( b ) 20 k ( c ) 50 k ( d ) 100 k ( e ) 150 k ( f ) 200 k ( g ) 250 k and ( h ) 300 k. figure 2(s2 ) ( a ) shows @xmath11 hysteresis loops of bi@xmath0gd@xmath1fe@xmath2ti@xmath3o@xmath4 ( x = 0.20 ) sample at different temperatures .
figure 2(s2 ) ( b - h ) present the enlarged view of the low field hysteresis loops at temperatures : ( b ) 20 k ( c ) 50 k ( d ) 100 k ( e ) 150 k ( f ) 200 k ( g ) 250 k and ( h ) 300 k. figure 3(s3 ) ( a - e ) show the m - h hysteresis loops at t = 150 k carried out at different cooling magnetic fields .
the m - h hysteresis loops at t = 250 k for different cooling magnetic fields are shown in figures 4(s4 ) ( a - e ) .
the asymmetric shifts toward the field axes are clearly visible from the enlarged views .
table 1 displays the variation of exchange bias fields , h@xmath33 of bi@xmath0gd@xmath1fe@xmath2ti@xmath3o@xmath4 ( x = 0.00 and x = 0.20 ) at different temperatures .
guo y , shi l , zhou s , zhao j , wang c , liu w and wei s 2013 _ j. phys .
_ d : _ appl .
phys . _ * 46 * 175302 zhai l , shi y g , tang s l , lv l y and du y w 2009 _ j. phys .
_ d : _ appl .
phys . _ * 42 * 165004 sati p c , arora m , chauhan s , chhoker s and kumar m 2012 _ j. appl .
phys . _ * 112 * 094102 suresh p and srinath s 2013 _ j. alloys and compounds _ * 554 * 271 das s r , choudhary r n p , bhattacharya p , katiyar r s , dutta p , manivannan a , and seehra m s 2007 _ j. app .
phys . _ * 101 * 034104 hussain s , hasanain s k , jaffari g h , ali n z , siddique m and shah s i 2015 _ j. alloys and compounds _ * 622 * 8 fertman e , dolya s , desnenko v , pozhar l a , kajnakova m and feher a 2014 _ j. app .
phys . _ * 115 * 203906 huang x h , ding j f , zhang g q , hou y , yao y p and li x g 2008 _ phys . rev .
_ b * 78 * 224408 qian j f , nayak a k , kreiner g , schnelle w and felser c 2014 _ j. phys .
_ d : _ appl .
phys . _ * 47 * 305001 park t j , papaefthymiou g c , viescas a j , lee y , zhou h and wong s s 2010 _ phys .
_ b * 82 * 024431 patankar s , pandey s k , reddy v r , gupta a , banerjee a and chaddah p 2010 _ epl _ * 90 * 57007 .
ruette b , zvyagin s , pyatakov a p , bush a , li j f , belotelov v i , zvezdin a k , and viehland d , 2004 _ phys .
_ b * 69 * 064114 fang l , liu j , ju s , zheng f , dong w and shen m 2010 _ appl .
* 97 * 242501 singh m k , prellier w , singh m p , katiyar r s , and scott j f 2008 _ phys .
_ b * 77 * 144403 rana d s , kuberkar d g and malik s k 2006 _ phys .
_ b * 73 * 064407 lu j , gnther a , schrettle f , mayr f , krohns s , lunkenheimer p , pimenov a , travkin v d , mukhin a a , and loid a , _ eur .
phys . j. _
b * 75 * 451 lazenka v v , zhang g , vanacken j , makoed i i , ravinski a f and moshchalkov v v 2012 _ j. phys .
d : _ appl .
phys . _ * 45 * 125002 murthy k j and venimadhav a 2013 _ appl .
_ * 103 * , 252410 . ramachandran b. , dixit a , naik r , lawes g , and ramachandra rao m s 2012 _ j. app .
_ * 111 * 023910 deng d , zheng j , yu d , wang b , sun d , avdeev m , feng z , jing c , lu b , ren w , cao s and zhang j 2015 _ app .
* 107 * 102404 dho j , qi x. , kim h , macmanus - driscoll j l , and blamire m g , adv .
18 , 1445 ( 2006 ) .
borisov p , hochstrat a , chen x , kleemann w , and binek c , phys .
94 , 117203 ( 2005 ) .
yao r , cao c , zheng c , and lei q , rsc adv .
3 , 24231 ( 2013 ) .
ambrosea t , and chien c l 1998 _ j. app .
phys . _ * 83 * 7222 ang r , sun y p , luo x , hao c y , zhu x b and song w h 2008 _ j. app .
phys . _ * 104 * 023914 huang s , shi l r , sun h g , zhu c m , tian z m and yuan s l 2014 _ appl .
lett . _ * 105 * 192904 karmakar s , taran s , bose e , chaudhuri b k , sun c p , huang c l and yang h d 2008 _ phys .
_ b * 77 * 144409 basith m a , ngo d -t , quader a , rahman m a , sinha b l , ahmmad b , hirose f and mlhave k 2014 _ nanoscale _ * 6 * 14336 | we have investigated the effect of temperature on magnetic properties of bi@xmath0gd@xmath1fe@xmath2ti@xmath3o@xmath4 ( x = 0.00 - 0.20 ) multiferroic system .
unexpectedly , the coercive fields ( @xmath5 ) of this multiferroic system increased with increasing temperature . the coercive fields and remanent magnetization were higher over a wide range of temperatures in sample x = 0.10 i.e. in sample having composition bi@xmath0gd@xmath1fe@xmath0ti@xmath1o@xmath4 than those of x = 0.00 and 0.20 compositions .
therefore , we have carried out temperature dependent magnetization experiments extensively for sample x = 0.10 .
the magnetic hysteresis loops at different temperatures exhibit an asymmetric shift towards the magnetic field axes which indicate the presence of exchange bias effect in this material system .
the hysteresis loops were also carried out at temperatures 150 k and 250 k by cooling down the sample from 300 k in various cooling magnetic fields ( @xmath6 ) . the exchange bias field ( @xmath7 )
values increased with @xmath6 and decreased with temperature .
the @xmath7 values were tunable by field cooling at temperatures up to 250 k. |
in this paper , we study problems connected with the renormalized electron mass in a model of quantum electrodynamics ( qed ) with nonrelativistic matter .
we are interested in rigorously controlling radiative corrections to the electron mass caused by the interaction of the electron with the _ soft _ modes of the quantized electromagnetic field .
the model describing interactions between nonrelativistic , quantum - mechanical charged matter and the quantized radiation field at low energies ( i.e. , energies smaller than the rest energy of an electron ) is the `` standard model '' , see @xcite . in this paper
, we consider a system consisting of a single spinless electron , described as a nonrelativistic particle that is minimally coupled to the quantized radiation field , and photons .
electron spin can easily be included in our description without substantial complications .
+ the physical system studied in this paper exhibits space translations invariance .
the hamiltonian , @xmath4 , generating the time evolution , commutes with the vector operator , @xmath1 , representing the total momentum of the system , which generates space translations .
if an infrared regularization , e.g. , an infrared cutoff @xmath5 on the photon frequency , is imposed on the interaction hamiltonian , there exist single - electron or dressed one - electron states , as long as their momentum is smaller than the bare electron mass , @xmath6 , of the electron .
this means that a notion of mass shell in the energy momentum spectrum is meaningful for velocities @xmath7 smaller than the speed of light @xmath8 ; ( with @xmath9 in our units ) .
vectors @xmath10 describing dressed one - electron states are normalizable vectors in the hilbert space @xmath11 of pure states of the system .
they are characterized as solutions of the equation @xmath12 where @xmath13 is the hamiltonian with an infrared cutoff @xmath5 in the interaction term and @xmath14 , the energy of a dressed electron , is a function of the momentum operator @xmath1 .
if in the joint spectrum of the components of @xmath1 the support of the vector @xmath15 is contained in a ball centered at the origin and of radius less than @xmath16 then eq . ( [ eq : i.1 ] ) has solutions ; see @xcite . since @xmath17=0 $ ] , eq .
( [ eq : i.1 ] ) can be studied for the fiber vectors , @xmath18 , corresponding to a value , @xmath1 , of the total momentum ; ( both the total momentum operator and points in its spectrum will henceforth be denoted by @xmath1 without danger of confusion ) .
thus we consider the equation @xmath19 where @xmath20 is the fiber hamiltonian at fixed total momentum @xmath1 , and @xmath21 is the value of the function @xmath22 at the point @xmath23 .
physically , states @xmath10 solving eq .
( [ eq : i.1 ] ) describe a freely moving electron in the absence of asymptotic photons . + it is an essential aspect of the `` infrared catastrophe '' in qed that eq .
( [ eq : i.1 ] ) does not have any normalizable solution in the limit where the infrared cut - off @xmath5 tends to zero , and the underlying dynamical picture of a freely moving electron breaks down ; see @xcite .
nevertheless , the limiting behavior of the function @xmath14 is of great interest for the following reasons . as long as @xmath24
, a natural definition of the renormalized electron mass , @xmath25 , is given by the formula @xmath26^{-1}\,.\ ] ] ( note that @xmath27 is invariant under rotations ) .
equation ( [ eq : i.2 ] ) is expected to remain meaningful in the limit @xmath28 . in particular , the quantity on the r.h.s .
( [ eq : i.2 ] ) is expected to be positive and bounded from above uniformly in the infrared cutoff @xmath5 .
more importantly , one aims at mathematical control of the function @xmath29^{-1}\ ] ] in a full neighborhood , @xmath30 , of @xmath2 , corresponding to a slowly moving electron ; ( i.e. , in the nonrelativistic regime ) .
when combined with a number of other spectral properties of the hamiltonian of nonrelativistic qed the condition @xmath31 uniformly in @xmath24 , suffices to yield a consistent scattering picture in the limit when @xmath28 in which the electron exhibits _ infraparticle _ behavior .
in fact , ( [ eq : i.5 ] ) is a crucial ingredient in the analysis of compton scattering presented in @xcite , @xcite .
+ _ main results _ assuming the coupling constant , @xmath32 , small enough , the following results follow .
\1 ) the function @xmath33 is well defined for @xmath34 ; furthermore , it is hlder - continuous in @xmath1 .
\2 ) the function @xmath35 is twice differentiable in @xmath36 and @xmath37 \3 ) @xmath38 uniformly in @xmath5 , where @xmath6 is the bare electron mass .
( our results can be extended to a region @xmath30 ( inside the unit ball ) of radius larger than @xmath39 . ) + we wish to mention some related earlier results . using _ operator - theoretic renormalization group methods _ , results ( [ eq : i.6 ] ) and ( [ eq : i.7 ] ) have been proven in @xcite for the special value @xmath2 .
the point @xmath2 is exceptional , because the hamiltonian @xmath40 is _ infrared regular _ at @xmath2 ; it has a normalizable ground state .
thomas chen ( see @xcite ) has established the results in ( [ eq : i.6bis ] ) , ( [ eq : i.7 ] ) ( using smooth infrared cut - offs ) by a highly non - trivial extension of the analysis of @xcite to arbitrary momenta @xmath36 .
the procedure presented in our paper relies on _ iterative analytic perturbation theory _ ( see section [ sectii ] where this tool is recalled ) that makes our proof substantially different and much shorter in comparison to a renormalization group approach .
the main feature is a more transparent treatment of the so called _ marginal _ terms of the interaction , where an essential role is played by explicit bogoliubov transformations that transform the infrared representations of the ccr of photon creation- and annihilation operators determined by dressed one - particle states of fixed momentum @xmath41 back to the fock representation .
the use of these bogoliubov transformations is a crucial device in our fight against the infrared problem .
the way in which we are using them is new , at least in the context of mathematically rigorous results on the infrared problem in qed . in our paper , the regularity properties of @xmath0 come with an explicit control of the asymptotics of the fiber ground state eigenvectors @xmath42 as @xmath5 tends to zero .
( this improves earlier results in @xcite . ) along the lines of @xcite , these results are preparatory to developing an infrared finite algorithm for the asymptotic expansion of the renormalized electron mass in powers and , probably , logarithms of the finestructure constant @xmath32 , up to an arbitrarily small remainder term .
however , the expansion in the coupling constant @xmath32 is not studied in this paper . with regard to _ ultraviolet _ corrections to the electron mass in nonrelativistic qed models ,
we refer the reader to @xcite , @xcite , @xcite , and @xcite . in section [ definition ] below ,
the model is defined rigorously .
then , for the convenience of the reader , in section [ outline ] we outline the key ideas of the proof and present the organization of the remaining sections of the paper . _ hilbert space _ the hilbert space of pure state vectors of a system consisting of one non - relativistic electron interacting with the quantized electromagnetic field is given by @xmath43 where @xmath44 is the hilbert space for a single schrdinger electron ; for expository convenience , we neglect the spin of the electron .
the hilbert space , @xmath45 , used to describe the states of the transverse modes of the quantized electromagnetic field ( the _ photons _ ) in the coulomb gauge is given by the fock space @xmath46 where @xmath47 is the vacuum vector ( the state of the electromagnetic field without any excited modes ) , and @xmath48 where the hilbert space @xmath49 of state vectors of a single photon is @xmath50 here , @xmath51 is momentum space , and @xmath52 accounts for the two independent transverse polarizations ( or helicities ) of a photon . in ( [ eq - i-3 ] ) , @xmath53 denotes the orthogonal projection onto the subspace of @xmath54 of totally symmetric @xmath55-photon wave functions , which accounts for the fact that photons satisfy bose - einstein statistics .
thus , @xmath56 is the subspace of @xmath45 of state vectors corresponding to configurations of exactly @xmath55 photons
. + _ units _ in this paper , we employ units such that planck s constant @xmath57 , the speed of light @xmath8 , and the mass of the electron @xmath6 are equal to 1 .
hamiltonian _
the dynamics of the system is generated by the hamiltonian @xmath58 the ( three - component ) multiplication operator @xmath59 represents the position of the electron .
the electron momentum operator is given by @xmath60 .
furthermore , @xmath61 is the fine structure constant ( which , in this paper , plays the rle of a small parameter ) , and @xmath62 denotes the vector potential of the transverse modes of the quantized electromagnetic field in the _ coulomb gauge _ , @xmath63 cutoff at high photon frequencies .
@xmath64 is the hamiltonian of the quantized , free electromagnetic field .
it is given by @xmath65 where @xmath66 and @xmath67 are the usual photon creation- and annihilation operators satisfying the canonical commutation relations @xmath68 & = & \delta_{\lambda \lambda ' } \ , \delta ( { { \vec{k}}}- { { \vec{k } } } ' ) { \ : , } \\
\label{eq - i-11 } [ a^\#_{{{\vec{k } } } , \lambda } \ , , \ , a^\#_{{{\vec{k } } } ' , \lambda ' } ] & = & 0\end{aligned}\ ] ] for @xmath69 and @xmath70 , where @xmath71 or @xmath72 .
the vacuum vector @xmath73 is characterized by the condition @xmath74 for all @xmath75 and @xmath76 .
the quantized electromagnetic vector potential is given by @xmath77 where @xmath78 , @xmath79 are photon polarization vectors , i.e. , two unit vectors in @xmath80 satisfying @xmath81 for @xmath82 . the equation
@xmath83 expresses the coulomb gauge condition .
moreover , @xmath84 is a ball of radius @xmath85 centered at the origin in momentum space .
here , @xmath85 represents an _ ultraviolet cutoff _ that will be kept fixed throughout our analysis .
the vector potential defined in ( [ eq - i-14 ] ) is thus cut off in the ultraviolet . throughout this paper
, it will be assumed that @xmath86 ( the rest energy of an electron ) , and that @xmath61 is sufficiently small . under these assumptions , the hamiltonian @xmath4 is selfadjoint on @xmath87 , i.e. , on the domain of definition of the operator @xmath88 the perturbation @xmath89 is small in the sense of kato .
the operator representing the total momentum of the system consisting of the electron and the electromagnetic radiation field is given by @xmath90 with @xmath91 , and where @xmath92 is the momentum operator associated with the photon field .
the operators @xmath4 and @xmath1 are essentially selfadjoint on a common domain , and since the dynamics is invariant under translations , they commute , @xmath17=\vec 0 $ ] .
the hilbert space @xmath11 can be decomposed into a direct integral over the joint spectrum , @xmath51 , of the three components of the momentum operator @xmath1 .
their spectral measure is absolutely continuous with respect to lebesgue measure , and hence we have that @xmath93 where each fiber space @xmath94 is a copy of fock space @xmath45 . + * remark * _ throughout this paper , the symbol @xmath1 stands for both a vector in @xmath51 and the vector operator on @xmath11 , representing the total momentum , depending on context .
similarly , a double meaning is given to arbitrary functions , @xmath95 , of the total momentum operator .
_ + we recall that vectors @xmath96 are given by sequences @xmath97 of functions , @xmath98 , where @xmath99 , of the electron position @xmath100 and of @xmath6 photon momenta @xmath101 and helicities @xmath102 , with the following properties : * @xmath103 is totally symmetric in its @xmath6 arguments @xmath104 .
* @xmath98 is square - integrable , for all @xmath6 . *
if @xmath105 and @xmath106 are two vectors in @xmath11 then @xmath107 we identify a square integrable function @xmath108 with the sequence @xmath97 where @xmath109 , and @xmath110 for all @xmath111 ; analogously , a square integrable function @xmath112 , @xmath113 , is identified with the sequence @xmath114 where @xmath115 , and @xmath116 for all @xmath117 . from now on , a sequence describing a quantum state with a fixed number of photons is identified with its nonzero component wave function ; vice versa , a wave function corresponds to a sequence according to the previous identification .
the elements of the fiber space @xmath118 are obtained by linear combinations of the ( improper ) eigenvectors of the total momentum operator @xmath1 with eigenvalue @xmath119 , e.g. , the plane wave @xmath120 is the eigenvector describing a state with an electron and no photon . given any @xmath121 , there is an isomorphism , @xmath122 , @xmath123 from the fiber space @xmath94 to the fock space @xmath124 , acted upon by the annihilation- and creation operators @xmath125 , @xmath126 , where @xmath125 corresponds to @xmath127 , and @xmath128 to @xmath129 , and with vacuum @xmath130 . to define @xmath122
more precisely , we consider a vector @xmath131 with a definite total momentum describing an electron and @xmath132 photons .
its wave function in the variables @xmath133 is given by @xmath134 where @xmath135 is totally symmetric in its @xmath132 arguments .
the isomorphism @xmath122 acts by way of @xmath136 because the hamiltonian @xmath4 commutes with the total momentum , it preserves the fibers @xmath94 for all @xmath121 , i.e. , it can be written as @xmath137 where @xmath138 written in terms of the operators @xmath125 , @xmath126 , and of the variable @xmath1 , the fiber hamiltonian @xmath40 is given by @xmath139 where @xmath140 and @xmath141 let @xmath142 in order to give a mathematically precise meaning to the constructions presented in the following , we introduce an infrared cut - off at a photon frequency @xmath24 in the vector potential .
the calculation of the second derivative of the energy of a dressed electron in the following called the `` ground state energy '' as a function of @xmath1 in the limit where @xmath28 , and for @xmath36 , represents the main problem solved in this paper .
hence we will , in the sequel , study the regularized fiber hamiltonian @xmath143 acting on the fiber space @xmath144 , for @xmath145 , where @xmath146 and where @xmath147 is a ball of radius @xmath5 centered at the origin . in the following , we will consider a sequence of infrared cutoffs @xmath148 with @xmath149 and @xmath150 .
+ _ notations _
\1 ) we use the notation @xmath151 for the norm of a bounded operator @xmath152 acting on a hilbert space @xmath11 .
typically , @xmath11 will be some subspace of @xmath153 .
\2 ) throughout the paper , we follow conventions such that @xmath154 where @xmath155 is an integration path in the complex space enclosing the origin . next , we outline the key ideas used in the proofs of our main results in eqs .
( [ eq : i.6 ] ) , ( [ eq : i.6bis ] ) , and ( [ eq : i.7 ] ) . for @xmath36 , @xmath32 small enough , and @xmath24
, @xmath14 is an isolated eigenvalue of @xmath156 ; see section ii and eq .
( [ eq : ii.4 ] ) .
because of the analyticity of @xmath157 in the variable @xmath1 , it follows that @xmath158\frac{1}{h^{\sigma}_{{{\vec{p}}}}-z}dz\,\psi_{{{\vec{p}}}}^{\sigma},\,[\partial_ih^{\sigma}_{{{\vec{p}}}}]\psi_{{{\vec{p}}}}^{\sigma}\rangle|_{{{\vec{p}}}=p^{i}\hat{i } } \ , , \nonumber\end{aligned}\ ] ]
where @xmath159 , @xmath160 is the unit vector in the direction @xmath161 , @xmath162 is the normalized ground state eigenvector of @xmath157 constructed in @xcite ; @xmath163 is a contour path in the complex energy plane enclosing @xmath14 and no other point of the spectrum of @xmath157 , and such that the distance of @xmath163 from @xmath164 is of order @xmath5 . at first glance , the expression on the r.h .
s. of ( [ eq : i.36 ] ) might become singular as @xmath28 , because the spectral gap above @xmath165 is of order @xmath5 . to prove that the limit @xmath28 is , in fact , well defined , we make use of a @xmath5-dependent bogoliubov transformation , @xmath166 ; ( see section ii , eq . ( [ eq : ii.3 ] ) ) . this transformation has already been employed in @xcite to analyze mass shell properties .
in fact , conjugation of @xmath20 by @xmath166 yields an infrared regularized hamiltonian @xmath167 with the property that the corresponding ground state , @xmath168 , has a non - zero limit , as @xmath28 . the hamiltonian @xmath169 has a `` canonical form '' derived in @xcite ( see also @xcite , where a similar operator has been used in the analysis of the nelson model ) : @xmath170 where @xmath171 is defined in eq .
( [ eq - iii.18 ] ) , @xmath172 is a c - number defined in eq .
( [ eq - ii.42 ] ) , and @xmath173 is a vector operator defined in eq .
( [ eq : ii.40 ] ) starting from eqs .
( [ eq : iii.16 ] ) , ( [ eq - iii.17 ] ) , ( [ eq : ii.34bis ] ) . by construction , @xmath174 this is a crucial property in the proof of existence of a limit of @xmath168 as @xmath28
( [ eq : i.39 ] ) is also an important ingredient in the proof of ( [ eq : i.6 ] ) , because , by applying the unitary operator @xmath166 to each term of the scalar product on the r.h.s . of ( [ eq : i.36 ] ) and using ( [ eq : i.39 ] ) ( see section [ sectiii ] ) , one finds that @xmath175\frac{1}{k^{\sigma_j}_{{{\vec{p}}}}-z_j}dz_j\,\phi_{{{\vec{p}}}}^{\sigma_j}\,,\nonumber\\ & & \quad\quad\quad\quad\quad\quad\quad\quad\quad\quad\,,\,[\partial_ie_{{{\vec{p}}}}^{\sigma_j}-(\gamma^{\sigma_j}_{{{\vec{p}}}})^{i}]\phi_{{{\vec{p}}}}^{\sigma_j}\rangle|_{{{\vec{p}}}=p^{i}\hat{i } } \label{eq : i.39bis } \\ & = & 1 - 2 \langle \frac{1}{2\pi i}\oint_{\gamma_{\sigma}}\frac{1}{k^{\sigma}_{{{\vec{p}}}}-z}\,(\gamma^{\sigma}_{{{\vec{p}}}})^i\ , \frac{1}{k^{\sigma}_{{{\vec{p}}}}-z}dz\,\frac{\phi_{{{\vec{p}}}}^{\sigma}}{\|\phi_{{{\vec{p}}}}^{\sigma}\|}\,,\,(\gamma^{\sigma}_{{{\vec{p}}}})^i\,\frac{\phi_{{{\vec{p}}}}^{\sigma}}{\|\phi_{{{\vec{p}}}}^{\sigma}\|}\rangle|_{{{\vec{p}}}=p^{i}\hat{i}}\ , .
\quad\quad\quad \label{eq : i.40}\end{aligned}\ ] ] notice that , if one starts from the two expressions on the r.h.s . of eq .
( [ eq : i.36 ] ) and on the r.h.s . of eq .
( [ eq : i.39bis ] ) respectively , both formally expanded in powers of @xmath176 , the bogoliubov transformation can be seen as a tool to re - collect an infinite number of terms and to show a nontrivial identity , thanks only to eq .
( [ eq : i.39 ] ) and to a vanishing contour integration ( see eq .
( [ eq : iii.48 ] ) ) .
next , still using eq .
( [ eq : i.39 ] ) , one can show that ( [ eq : i.40 ] ) remains uniformly bounded in @xmath5 .
+ to see this we use the inequality @xmath177 for an arbitrarily small @xmath178 , with @xmath179 and @xmath32 small enough depending on @xmath180 . this inequality will be proven inductively ( see theorem [ th.induction ] ) by introducing sequences of infrared cut - offs @xmath181 , where @xmath182 as @xmath183 .
the proof by induction is combined with an improved ( as compared to the result in @xcite ) estimate of the rate of convergence of @xmath184 as @xmath185 . by telescoping , one can plug these improved estimates into ( [ eq : i.40 ] ) to end up with the desired uniform bound .
the control of the rate of convergence of the r.h.s . in ( [ eq : i.40 ] ) , as @xmath28 , combined with the smoothness in @xmath1 , for arbitrary infrared cutoff @xmath24 , finally entails the hlder - continuity in @xmath1 of the limiting quantity @xmath186 the hlder - continuity in @xmath1 of @xmath187 and of @xmath188 , combined with the fundamental theorem of calculus , imply that @xmath0 is twice differentiable and @xmath189 . + our paper is organized as follows . in section [ sectii ]
, we recall how to construct the ground states of the hamiltonians @xmath157 and @xmath169 by _ iterative analytic perturbation theory_. this section contains an explicit derivation of the formula of the transformed hamiltonians and of related algebraic identities that will be used later on . in section [ sectiii ] , we first derive inequality ( [ eq : i.41 ] ) and the improved convergence rate of @xmath184 as @xmath28 , by using some key ingredients described in section [ sectii ] .
section [ sectiii.1 ] is devoted to an analysis of ( [ eq : i.36 ] ) that culminates in the following main results .
_ * theorem * _ _ for @xmath32 small enough , @xmath190 converges as @xmath28 .
the limiting function @xmath191 is hlder continuous in @xmath36 .
the limit @xmath192 holds true uniformly in @xmath36 .
_ _ * corollary * _ _ for @xmath32 small enough , the function @xmath193 , @xmath36 , is twice differentiable , and @xmath194 _ + *
remark * for the complete proof of the construction of the ground states of the hamiltonians @xmath157 and @xmath169 by _ iterative analytic perturbation theory _ , the reader is advised to consult ref . @xcite .
in this section , we report on results contained in @xcite concerning the ground states of the hamiltonians @xmath195 , where @xmath36 and @xmath196 , and the existence of a limiting vector for the sequence of ground state vectors of the transformed hamiltonians , @xmath197 , where the bogoliubov transformation used to obtain @xmath197 from @xmath195 ( derived in @xcite ) is determined by @xmath198 for @xmath199 , with @xmath200 where @xmath201 is defined in ( [ eq - iii.18 ] ) . in @xcite ,
the first step consists in constructing the ground states of the regularized fiber hamiltonians @xmath202 . as shown in @xcite
, @xmath203 has a unique ground state , @xmath204 , that can be constructed by _ iterative analytic perturbation theory _ , as developed in @xcite . to recall how this method works some preliminary definitions and results
are needed : * we introduce the fock spaces @xmath205 note that @xmath206 if not specified otherwise , @xmath207 denotes the vacuum state in anyone of these fock spaces .
any vector @xmath208 in @xmath209 can be identified with the corresponding vector , @xmath210 , in @xmath45 , where @xmath207 is the vacuum in @xmath211 . *
momentum - slice interaction hamiltonians are defined by @xmath212 where @xmath213 * four real parameters , @xmath214 , and @xmath215 , will appear in our analysis .
they have the properties @xmath216 where @xmath217 , with @xmath218 , for @xmath32 small enough , is a constant such that the inequality @xmath219 holds for all @xmath36 and any @xmath220 . here
we note that @xmath222 , as @xmath223 ; ( see statement ( @xmath224 ) of theorem 3.1 . in @xcite ) . by iterative analytic perturbation theory ( see @xcite ) ,
one derives the following results , valid for sufficiently small @xmath32 , depending on our choice of @xmath225 and @xmath226 ( see also figure 1 below ) : * @xmath227 is an isolated simple eigenvalue of @xmath228 with spectral gap larger or equal to @xmath229 .
furthermore , @xmath227 is also the ground state energy of @xmath230 , and it is an isolated simple eigenvalue of @xmath230 with spectral gap larger or equal to @xmath231 . * the ground - state energies @xmath227 and @xmath232 of the hamiltonians @xmath195 and @xmath233 , respectively , ( acting on the same space @xmath234 ) satisfy the inequalities @xmath235 where @xmath8 is independent of @xmath236 and of @xmath32 . *
the ground state vectors , @xmath237 , of @xmath233 can be recursively constructed starting from @xmath238 with the help of the spectral projection @xmath239 more precisely , @xmath240^n\,\psi_{{{\vec{p}}}}^{\sigma_{j}}\otimes\omega_f \,,\quad\quad\quad\,\,\label{eq : ii.13}\end{aligned}\ ] ] where @xmath241 , with @xmath215 as in ( [ eq : iii.8bis ] ) .
@xmath237 is the ( unnormalized ) ground state vector of @xmath242 for any @xmath243 . in this section
, we consider the ( bogoliubov - transformed ) hamiltonians @xmath244 with ground state vectors @xmath245 , @xmath246 . some algebraic manipulations to express @xmath197 in a
canonical form " appear to represent a crucial step before iterative perturbation theory can be applied to the sequence of these transformed hamiltonians .
in addition , some intermediate hamiltonians , denoted @xmath247 , must be introduced to arrive at the right kind of convergence estimates .
the same algebraic relations that are used to obtain the canonical form " of @xmath197 also play an important role in the proof of our main result concerning the limiting behavior , as @xmath248 , of the second derivative of the ground state energy @xmath21 .
it is therefore useful to derive the canonical form " of @xmath197 and the relevant algebraic identities in some detail .
+ the feynman - hellman formula ( which holds because @xmath249 is an analytic family of type a , and @xmath227 is an isolated eigenvalue ) yields the identity @xmath250 where , given an operator @xmath251 and a vector @xmath252 in the domain of @xmath251 , we use the notation @xmath253 as stated in @xcite , for @xmath32 small enough , @xmath254 we define @xmath255 we rewrite @xmath195 as @xmath256 and , using ( [ eq - iii.15 ] ) and ( [ eq - iii.17 ] ) , @xmath257 we then obtain @xmath258 adding and subtracting @xmath259 , one finds that @xmath260 next , we implement the bogoliubov transformation @xmath261 for @xmath199 , where @xmath262 is defined in ( [ eq : ii.3 ] ) .
it is evident that @xmath263 acts as the identity on @xmath264 and on @xmath265 .
we define the vector operators @xmath266 noting that , by ( [ eq - iii.15 ] ) , ( [ eq - iii.17 ] ) , and ( [ eq : ii.34bis ] ) @xmath267 where @xmath245 is the ground state of the bogoliubov - transformed hamiltonian @xmath268 it is easy to see that @xmath269 after inspecting straightforward operator domain questions ( see @xcite ) , the canonical form " of @xmath197 is given by @xmath270 where @xmath271 so that @xmath272 and @xmath273 eqs .
( [ eq : iii.38 ] ) , ( [ eq : iii.39 ] ) follow by using that @xmath274 for @xmath275 . +
an intermediate hamiltonian , @xmath276 , is defined by @xmath277 where @xmath278
we decompose @xmath276 into several different terms , similarly as @xmath197 .
we recall that @xmath279 and , by ( [ eq : iii.35bis ] ) , @xmath257 it follows that ( see also ( [ eq : ii.28bis])-([eq : ii.31bis ] ) ) @xmath280 we now add and subtract @xmath281 and conjugate the resulting operator with the unitary operator @xmath282 . after inspecting straightforward operator domain questions ( see @xcite ) , we find that @xmath283 where @xmath284\frac{d^3k}{\sqrt{|{{\vec{k}}}|}}\nonumber\\ \hat{{{\mathcal{e}}}}_{{{\vec{p}}}}^{\sigma_{j+1}}&:= & \frac{{{\vec{p}}}^2}{2}-\frac{({{\vec{p}}}-{{\vec{\nabla}}}e_{{{\vec{p}}}}^{\sigma_j})^2}{2}\\ & & -\alpha\sum_{\lambda}\int_{{{\mathcal{b}}}_{\lambda}\setminus{{\mathcal{b}}}_{\sigma_{j+1}}}|{{\vec{k}}}|\delta_{{{\vec{p}}}}^{\sigma_j}(\hat{k})\,\frac{{{\vec{\nabla}}}e_{{{\vec{p}}}}^{\sigma_j}\cdot\vec{\epsilon}_{{{\vec{k}}},\lambda}}{|{{\vec{k}}}|^{\frac{3}{2}}\delta_{{{\vec{p}}}}^{\sigma_j}(\hat{k})}\frac{{{\vec{\nabla}}}e_{{{\vec{p}}}}^{\sigma_j}\cdot\vec{\epsilon}^{\,\,*}_{{{\vec{k}}},\lambda}}{|{{\vec{k}}}|^{\frac{3}{2}}\delta_{{{\vec{p}}}}^{\sigma_j}(\hat{k})}d^3k\,.\nonumber\end{aligned}\ ] ] we also define the operators @xmath285 and @xmath286 which are used in the proofs of convergence of the ground state vectors . here , @xmath287 denotes the ground state vector of the hamiltonian @xmath288 notice that @xmath289 an important identity used in @xcite and in the sequel of the present paper is ( @xmath290 ) @xmath291\frac{d^3k}{\sqrt{|{{\vec{k}}}|}}\,.\nonumber\end{aligned}\ ] ] eq .
( [ eq : iii.63 ] ) can be derived using ( [ eq : ii.34bis ] ) , ( [ eq : iii.36bis ] ) , ( [ eq : iii.38 ] ) , ( [ eq : ii.40 ] ) , ( [ eq : iii.61 ] ) , and ( [ eq : iii.62 ] ) . to pass from momentum scale @xmath236 to @xmath293
, we proceed in two steps : first , we construct an intermediate vector , @xmath294 , defined by @xmath295^n\phi_{{{\vec{p}}}}^{\sigma_{j}}\,,\ ] ] where @xmath296 subsequently , we construct @xmath297 by setting @xmath298 the series in ( [ eq : iii.64 ] ) is term - wise well - defined and converges strongly to a non - zero vector , provided @xmath32 is small enough ( _ independently _ of @xmath236 ) .
the proof of this claim is based on operator - norm estimates of the type used in controlling the neumann expansion in ( [ eq : ii.13 ] ) , which requires an estimate of the spectral gap that follows from the unitarity of @xmath299 and result 1 ) described after eq .
( [ eq : ii.10 ] ) .
a key point in our proof of convergence of the sequence @xmath300 is to show that the term @xmath301 appearing in ( [ eq : iii.53 ] ) , which is superficially marginal " in the infrared , by power counting , is in fact irrelevant " ( using the terminology of renormalization group theory ) .
this is a consequence of the orthogonality condition @xmath302 which , when combined with an inductive argument , implies that @xmath303\,(\frac{1}{k_{{{\vec{p}}}}^{\sigma_j}-z_{j+1}})^{\frac{1}{2}}\,\phi_{{{\vec{p}}}}^{\sigma_{j}}\|\ ] ] ( where @xmath304 stands for the part which contains only photon creation operators ) is of order @xmath305 , for some @xmath306 specified in @xcite .
in particular , this suffices to show that @xmath307 for any @xmath308 provided @xmath32 is sufficiently small . finally , in theorem 3.1 of ref .
@xcite , it is proven that there is a non - zero vector in the hilbert space corresponding to @xmath309 , and that the rate of convergence is at least @xmath310 for any @xmath308 provided @xmath32 is sufficiently small . + * remark * in theorem 3.1 of ref .
@xcite , for @xmath309 , the range of values of @xmath32 such that the rate of convergence , @xmath310 , holds is not claimed to be uniform in @xmath180 .
the stronger result obtained in the next section ( see ( [ eq : iii.bis.2 ] ) and ( [ eq : iii.34bis ] ) ) implies that this range ( corresponding to the rate @xmath310 ) is actually @xmath180-independent . to prove convergence of the sequence @xmath300 of ground state vectors of the hamiltonians @xmath197 , some further conditions on @xmath32 , @xmath311 , and @xmath215 ( see ( [ eq : iii.8bis ] ) , ( [ eq : iii.9 ] ) )
are required , in particular an upper bound on @xmath215 and an upper bound on @xmath311 strictly smaller than the ones imposed by inequalities ( [ eq : iii.8bis ] ) , ( [ eq : iii.9 ] ) ; ( for details , see lemma a.3 in @xcite ) .
we note that the more restrictive conditions on @xmath215 and @xmath311 imply new bounds on @xmath312 and @xmath313 .
moreover , @xmath311 must satisfy a lower bound @xmath314 , with a multiplicative constant @xmath315 sufficiently large .
some key inequalities needed in our analysis of the convergence properties of @xmath300 are summarized below .
they will be marked by the symbol @xmath316 . in order to reach some important improvements in our estimates of the convergence rate of @xmath317 , as @xmath318 ( discussed in the next section ) , a refined estimate
is needed that is stated in @xmath319 , and a new inequality , see @xmath320 , ( analogous to @xmath321 and @xmath322 ) is required .
* _ estimates on the shift of the ground state energy and its gradient _ + there are constants @xmath323 , @xmath324 such that the following inequalities hold .
+ @xmath325 @xmath326 see @xcite . + * @xmath319 @xmath327 , @xmath328 , and following the steps in the proof of lemma a.2 in @xcite .
* _ bounds relating expectations of operators to expectations of their absolute values _ + there are constants @xmath329 such that the following inequalities hold .
+ @xmath321 for @xmath330 , @xmath331 where @xmath332 is the @xmath333 component of @xmath334 .
+ @xmath322 for @xmath335 @xmath336 where @xmath337 is the @xmath338 component of @xmath304 .
+ @xmath320 for @xmath335 @xmath339 to prove @xmath321 and @xmath322 , it suffices to exploit the fact that the spectral support ( with respect to @xmath197 ) of the two vectors @xmath340 and @xmath341 is strictly above the ground state energy of @xmath342 , since they are both orthogonal to the ground state , @xmath317 , of this operator . in the proof of bound @xmath320
, it is also required that @xmath343 , as will be assumed in the following . + _ * remarks * _ \(1 ) the constants @xmath344 are independent of @xmath345 , and @xmath346 , provided that @xmath347 , and @xmath348 are sufficiently small .
( 2 ) for the convenience of the reader , we recapitulate the relations between the parameters entering the construction : @xmath349 moreover , we stress that the final result is a small coupling result , i.e. , valid for small values of @xmath32 , and that , for technical reasons , small values of the parameters @xmath350 within the constraints listed above ( that imply more restrictive bounds on @xmath351 ) are required .
+ the crucial estimate for the bound on @xmath352 obtained in @xcite ( see ( [ eq : iii.58 ] ) ) is @xmath353 where @xmath354 is independent of @xmath236 , and @xmath180 , @xmath308 , can be taken arbitrarily small for @xmath32 and @xmath355 sufficiently small ( depending on @xmath180 ) . this estimate will be improved in the next section . as a consequence , our estimate of the convergence rate of @xmath356 will be improved . as a corollary ,
the second derivative of @xmath14 is proven to converge , as @xmath28 .
our arguments in section iii rely on the results previously proven in @xcite and described in section ii , which hold for @xmath32 small enough . therefore ,
in the following , we assume the constraints ( [ eq : ii.79bis])-([eq : ii.82bisbis ] ) , and we make use of the estimates on the spectral gaps ( see @xmath358 in section [ sectii.0 ] ) and of the bounds @xmath325-@xmath320 ( see section [ sectii.2 ] ) .
we also make use of the lower bounds @xmath359 uniformly in @xmath360 , which appear in the proof of theorem 3.1 of ref .
@xcite . assuming these bounds we can simplify the proof by induction in the theorem below .
+ [ th.induction ] for @xmath347 sufficiently small ( depending on @xmath180 ) , the inequality @xmath361 holds true where @xmath308 and @xmath362 is a constant independent of @xmath363 .
furthermore , for @xmath362 and @xmath32 small enough , inequality ( [ eq : iv.1 ] ) implies @xmath364 * _ _ + we assume that , at scale @xmath365 , the following estimate holds @xmath366 this estimate readily implies that , for @xmath362 and @xmath367 small enough , but uniformly in @xmath236 , @xmath368^n\phi_{{{\vec{p}}}}^{\sigma_{j-1}}\|\,\nonumber\\ & \leq&\alpha^{\frac{1}{4}}\epsilon^{j(1-\delta)}\,.\label{eq : iii.71}\end{aligned}\ ] ] an improved estimate on @xmath369 is based on the following bounds : * * @xmath370 + whose proof requires the use of the pull - through formula " ( see , e.g. , @xcite ) , a neumann expansion of the resolvent , the inequality in eq .
( [ eq : ii.78 ] ) , and eq .
( [ eq : iv.3 ] ) ; the reader can follow the similar steps used in lemma a3 of ref .
@xcite ; * * @xmath371 + this estimate can be derived from standard bounds and using the pull - through formula " . *
_ _ + by unitarity of @xmath372 , we have that @xmath373 for @xmath32 small enough and @xmath374 , where @xmath315 is large enough , we may use @xmath325 to re - expand the resolvent and find that @xmath375 it follows that @xmath376 our recursion , combined with ( [ eq : ii.78 ] ) , relates ( [ eq : iii.80 ] ) to the initial expression in ( [ eq : iv.1 ] ) , with @xmath236 replaced by @xmath377 , while ( [ eq : iii.79 ] ) is a remainder term .
next we note that @xmath378 here @xmath379 and @xmath380 are constants independent of @xmath32 , @xmath215 , and @xmath346 , provided that @xmath32 , @xmath215 are sufficiently small , and @xmath381 . in detail : * * property ( @xmath382 ) and the two norm - bounds @xmath383 ( see ( [ eq : iii.70 ] ) ) justify the step from ( [ eq : iii.83 ] ) to ( [ eq : iii.84 ] ) ; * * concerning the step from ( [ eq : iii.82 ] ) to ( [ eq : iii.84 ] ) , it is enough to consider eq .
( [ eq : iii.63 ] ) and the two bounds @xmath384 ( hint : for the first inequality in ( [ eq : iii.20b ] ) , use the expression in ( [ eq : iii.64 ] ) . )
+ to bound the term ( [ eq : iii.80 ] ) , we use @xmath320 and the key orthogonality property ( [ eq : iii.55 ] ) . for @xmath385 and @xmath330 , we find that for @xmath386 sufficiently small @xmath387 in passing from ( [ eq : iii.87 ] ) to ( [ eq : iii.88 ] ) , we again use the constraint on the spectral support ( with respect to @xmath388 ) of the vector @xmath389 .
+ assuming that the parameters @xmath311 and @xmath32 are so small that the previous constraints are fulfilled and that @xmath390 we then conclude that @xmath391 + notice that the bound in ( [ eq : iii.89 ] ) induces a @xmath392dependent constraint on the admissible values of @xmath311 and , due to ( [ eq : ii.83b ] ) , on @xmath32 . *
_ _ + since @xmath393 inequality ( [ eq : iv.1 ] ) is trivially fulfilled for @xmath394 ; thus ( [ eq : iv.1 ] ) holds for all @xmath360 and for @xmath362 _ arbitrarily _ small , provided @xmath32 is small enough . as we explain below , an improved estimate of the rate of convergence of the sequence @xmath395 follows from the bound in ( [ eq : iii.bis.2 ] ) , but we stress that only the estimates in eqs .
( [ eq : iv.1 ] ) , ( [ eq : iii.bis.2 ] ) will be used for the uniform bound on the second derivative of @xmath14 in next section .
in fact , one can combine the bound in eq .
( [ eq : iii.bis.2 ] ) with the estimate @xmath396 where @xmath397 is independent of @xmath32 , @xmath311 , @xmath215 , and @xmath346 , provided that @xmath32 , @xmath311 , and @xmath215 are sufficiently small .
the estimate in ( [ eq : iii.30bis ] ) is obtained starting from the definition in eq .
( [ eq : ii.bis.67 ] ) and using the soft photon bound @xmath398 that follows from inequality ( [ eq : ii.10 ] ) and the identity @xmath399 which is derived in @xcite by using a pull - through argument " . by a standard procedure ( see , e.g. , @xcite ) , one obtains similar results for the ground state vectors of the @xmath5-dependent hamiltonians @xmath169 , for arbitrary @xmath24 .
a precise statement concerning the rate of convergence is as follows : the _ normalized _ ground state vectors ( that , with an abuse of notation , we call @xmath400 ) @xmath401 where @xmath402 , converge strongly to a vector @xmath403 , as @xmath28 , with @xmath404 for any @xmath405 , provided @xmath32 is in an interval @xmath406 where @xmath407 , @xmath408 as @xmath409 .
the @xmath180-dependence of the interval @xmath410 is an indirect consequence of the upper bound on @xmath311 that must be imposed through ( [ eq : iii.89 ] ) to implement the proof by induction .
the relations ( [ eq : ii.79bis])-([eq : ii.82bisbis ] ) induce a @xmath180-dependence on the other parameters and in particular on @xmath32 .
another @xmath180-type dependence of the estimated rate of convergence , @xmath411 , comes from the logarithmic term in ( [ eq : iii.30bis ] ) .
however , this does not spoil the uniformity in @xmath180 of the interval of admissible values of @xmath32 but only affects the multiplicative constant on the r.h.s . of ( [ eq : iii.34bis ] ) .
moreover , with further work , the estimate in eq .
( [ eq : iii.30bis ] ) can be improved to remove the logarithmic term .
because of rotational symmetry we have that @xmath412 .
moreover , @xmath413 is an analytic family of type a in @xmath36 , with an isolated eigenvalue @xmath414 .
thus , the second derivative @xmath415 is well defined and @xmath416 where @xmath417 . without loss of generality ,
the following results are proven for the standard sequence @xmath418 of infrared cutoffs . by simple arguments ( see @xcite ) , limiting behavior as @xmath28 is shown to be sequence - independent " . by analytic perturbation
theory we have that @xmath419\frac{1}{h^{\sigma_j}_{{{\vec{p}}}}-z_j}dz_j\,\psi_{{{\vec{p}}}}^{\sigma_j},\,[p^{i}-(\beta^{\sigma_j})^{i}]\psi_{{{\vec{p}}}}^{\sigma_j}\rangle|_{{{\vec{p}}}=p^{i}\hat{i } } \ , , \nonumber\end{aligned}\ ] ] here @xmath204 is the normalized ground state eigenvector of @xmath195 .
next , we make use of the bogoliubov transformation implemented by @xmath262 to show that @xmath420\frac{1}{h^{\sigma_j}_{{{\vec{p}}}}-z_j}dz_j\,\psi_{{{\vec{p}}}}^{\sigma_j},\,[p^{i}-(\beta^{\sigma_j})^{i}]\psi_{{{\vec{p}}}}^{\sigma_j}\rangle \\ & = & \frac{1}{\|\phi_{{{\vec{p}}}}^{\sigma_j}\|^2}\langle \frac{1}{2\pi i}\oint_{\gamma_j}\frac{1}{k^{\sigma_j}_{{{\vec{p}}}}-z_j}[p^{i}-w_{\sigma_j}({{\vec{\nabla}}}e_{{{\vec{p}}}}^{\sigma_j})(\beta^{\sigma_j})^{i}w_{\sigma_j}^*({{\vec{\nabla}}}e_{{{\vec{p}}}}^{\sigma_j})]\frac{1}{k^{\sigma_j}_{{{\vec{p}}}}-z_j}dz_j\,\phi_{{{\vec{p}}}}^{\sigma_j}\,,\nonumber\\ & & \quad\quad\quad\quad\quad,\,[p^{i}-w_{\sigma_j}({{\vec{\nabla}}}e_{{{\vec{p}}}}^{\sigma_j})(\beta^{\sigma_j})^{i}w_{\sigma_j}^*({{\vec{\nabla}}}e_{{{\vec{p}}}}^{\sigma_j})]\phi_{{{\vec{p}}}}^{\sigma_j}\rangle \,,\end{aligned}\ ] ] where @xmath245 is the ground state eigenvector of @xmath197 ( iteratively constructed in section ii ) .
recalling the definitions @xmath421 @xmath422 and because of the identity ( feynman - hellman , see ( [ eq : iii.36bis ] ) ) @xmath423 we find that @xmath424 hence , @xmath425\frac{1}{k^{\sigma_j}_{{{\vec{p}}}}-z_j}dz_j\,\phi_{{{\vec{p}}}}^{\sigma_j}\,,\nonumber\\ & & \quad\quad\quad\quad\quad\quad\quad\quad\quad\quad\,,\,[\partial_ie_{{{\vec{p}}}}^{\sigma_j}-(\gamma^{\sigma_j}_{{{\vec{p}}}})^{i}]\phi_{{{\vec{p}}}}^{\sigma_j}\rangle|_{{{\vec{p}}}=p^{i}\hat{i } } \,.\end{aligned}\ ] ] using the eigenvalue equation @xmath426 the terms proportional to @xmath427 and to the mixed terms
i.e. , proportional to the product of @xmath428 and @xmath429 are seen to be identically @xmath430 , because the contour integral vanishes for each @xmath431 ; e.g. , @xmath432\frac{1}{k^{\sigma_j}_{{{\vec{p}}}}-z_j}\,\phi_{{{\vec{p}}}}^{\sigma_j},\,[\partial_ie_{{{\vec{p}}}}^{\sigma_j}]\phi_{{{\vec{p}}}}^{\sigma_j}\rangle d\bar{z}_j\quad\quad}\quad\\ & = & \oint_{\gamma_j}\langle\phi_{{{\vec{p}}}}^{\sigma_j}\,,\,\phi_{{{\vec{p}}}}^{\sigma_j}\rangle \big(\frac{\partial_ie_{{{\vec{p}}}}^{\sigma_j}}{e^{\sigma_j}_{{{\vec{p}}}}-\bar{z_j}}\big)^2d\bar{z}_j=0\,.\nonumber\end{aligned}\ ] ] by unitarity of @xmath372 , @xmath437 by assumption , @xmath32 is so small that the neumann series expansions of the resolvents below converge in @xmath438 : @xmath439 @xmath440 where : @xmath441^l\,,\nonumber\end{aligned}\ ] ] and @xmath442 is defined in eq .
( [ eq : iii.53 ] ) ; @xmath443 we proceed by using the obvious identity : @xmath444 each of the expressions ( [ eq : iv.57 ] ) and ( [ eq : iv.58 ] ) can be rewritten by adding and subtracting @xmath445 . for ( [ eq : iv.57 ] )
we get @xmath446\big \rangle\frac{1}{e^{\sigma_{j-1}}_{{{\vec{p}}}}-\bar{z}_{j } } d\bar{z}_{j}\nonumber\\ & & + \oint_{\gamma_{j}}\big\langle ( \hat{\gamma}_{{{\vec{p}}}}^{\sigma_j})^{i}\frac{\hat{\phi}_{{{\vec{p}}}}^{\sigma_{j}}}{\|\hat{\phi}_{{{\vec{p}}}}^{\sigma_{j } } \|}-(\gamma_{{{\vec{p}}}}^{\sigma_{j-1}})^{i}\frac{\phi_{{{\vec{p}}}}^{\sigma_{j-1}}}{\|\phi_{{{\vec{p}}}}^{\sigma_{j-1}}\|}\,,\label{eq : iv.63}\\ & & \quad\quad\quad,\,\frac{1}{k_{{{\vec{p}}}}^{\sigma_{j-1}}-\bar{z}_{j}}(\gamma_{{{\vec{p}}}}^{\sigma_{j-1}})^{i}\frac{\phi_{{{\vec{p}}}}^{\sigma_{j-1}}}{\|\phi_{{{\vec{p}}}}^{\sigma_{j-1}}\|}\big \rangle\frac{1}{e^{\sigma_{j-1}}_{{{\vec{p}}}}-\bar{z}_{j } } d\bar{z}_{j}\,\nonumber\\ & & + \oint_{\gamma_{j}}\big\langle ( \gamma_{{{\vec{p}}}}^{\sigma_{j-1}})^{i}\frac{\phi_{{{\vec{p}}}}^{\sigma_{j-1}}}{\|\phi_{{{\vec{p}}}}^{\sigma_{j-1}}\|}\,,\label{eq : iv.64}\\ & & \quad\quad\quad,\,\frac{1}{k_{{{\vec{p}}}}^{\sigma_{j-1}}-\bar{z}_{j}}[(\hat{\gamma}_{{{\vec{p}}}}^{\sigma_j})^{i}\frac{\hat{\phi}_{{{\vec{p}}}}^{\sigma_{j}}}{\|\hat{\phi}_{{{\vec{p}}}}^{\sigma_{j}}\|}-(\gamma_{{{\vec{p}}}}^{\sigma_{j-1}})^{i}\frac{\phi_{{{\vec{p}}}}^{\sigma_{j-1}}}{\|\phi_{{{\vec{p}}}}^{\sigma_{j-1}}\|}]\big \rangle\frac{1}{e^{\sigma_{j-1}}_{{{\vec{p}}}}-\bar{z}_{j } } d\bar{z}_{j}\,\nonumber\,.\end{aligned}\ ] ] the difference in eq .
( [ eq : iv.49 ] ) corresponds to the sum of the terms ( [ eq : iv.58])-([eq : iv.59 ] ) and of the terms ( [ eq : iv.62])-([eq : iv.64 ] ) .
in fact , ( [ eq : iv.61 ] ) corresponds to the first term in ( [ eq : iv.49 ] ) after a contour deformation from @xmath447 to @xmath448 .
the sum of the remainder terms ( [ eq : iv.58 ] ) , ( [ eq : iv.59 ] ) , and ( [ eq : iv.62])-([eq : iv.64 ] ) can be bounded by @xmath449 , for @xmath362 and @xmath32 small enough but independent of @xmath236 , for any @xmath36 .
( we recall that @xmath362 can be taken arbitrarily small , provided @xmath32 is small enough ) .
the details are as follows .
* for ( [ eq : iv.62])-([eq : iv.64 ] ) use the following inequalities @xmath450\,\hat{\phi}_{{{\vec{p}}}}^{\sigma_{j}}\big\| \leq { { \mathcal{o}}}(\alpha^{\frac{1}{4 } } \epsilon^{j(1-\delta)})\,,\label{eq : iv.66}\\ & & \big\|\frac{1}{k_{{{\vec{p}}}}^{\sigma_{j-1}}-\bar{z}_{j } } \big\|_{{{\mathcal{f}}}_{\sigma_j } } \leq { { \mathcal{o}}}(\frac{1}{\epsilon^{j}})\,,\\ & & \big\|\big(\frac{1}{k_{{{\vec{p}}}}^{\sigma_{j-1}}-\bar{z}_{j}}\big)(\gamma_{{{\vec{p}}}}^{\sigma_{j-1}})^{i}\,(\hat{\phi}_{{{\vec{p}}}}^{\sigma_{j}}-\phi_{{{\vec{p}}}}^{\sigma_{j-1}})\big\| \leq { { \mathcal{o}}}(\frac{\alpha^{\frac{1}{4 } } \epsilon^{j(1-\delta)}}{\epsilon^{j}})\,,\\ & & \big\|\hat{\phi}_{{{\vec{p}}}}^{\sigma_{j}}-\phi_{{{\vec{p}}}}^{\sigma_{j-1}}\big\| \leq \alpha^{\frac{1}{4 } } \epsilon^{j(1-\delta)}\,.\label{eq : iv.68}\end{aligned}\ ] ] in order to derive the inequality in eq . ( [ eq : iv.66 ] ) , one uses eqs .
( [ eq : iii.63 ] ) , ( [ eq : iii.73 ] ) , and ( [ eq : iii.57])-(ii.58 ) . * for ( [ eq : iv.58 ] ) , after adding and subtracting @xmath445 , one also has to use that @xmath451\frac{1}{k_{{{\vec{p}}}}^{\sigma_{j-1}}-z_{j}}(\gamma_{{{\vec{p}}}}^{\sigma_{j-1}})^{i}\phi_{{{\vec{p}}}}^{\sigma_{j-1}}\|\leq { { \mathcal{o}}}(\alpha^{\frac{1}{2}}\epsilon^{j-1}\,\frac{{{\mathcal{r}}}_0^{\frac{1}{4}}}{\alpha^{\frac{1}{4}}\ , \epsilon^{(j-1)\delta}})\,;\ ] ] * to bound ( [ eq : iv.59 ] ) , note that @xmath452 where @xmath453 . then use that @xmath454 and the following inequality analogous to ( [ eq : iv.65 ] ) @xmath455 for @xmath32 small enough
, @xmath190 converges , as @xmath28 .
the limiting function , @xmath456 , is hlder - continuous in @xmath36 ( for an exponent @xmath457 ) .
the limit @xmath458 holds true uniformly in @xmath36 .
we now observe that @xmath461 ( see eq .
( [ eq : iv.48 ] ) ) , because @xmath462 according to the constraint in eq .
( [ eq : ii.83b ] ) , we can take @xmath463 so that , for @xmath32 small enough , lemma [ lemmaiii.1 ] and ( [ eq : iv.72 ] ) yield @xmath464 uniformly in @xmath346 . hence the limit ( [ eq : iv.71 ] ) follows . | within the framework of nonrelativistic qed , we prove that , for small values of the coupling constant , the energy function , @xmath0 , of a dressed electron is twice differentiable in the momentum @xmath1 in a neighborhood of @xmath2 .
furthermore , @xmath3 is bounded from below by a constant larger than zero .
our results are proven with the help of _ iterative analytic perturbation theory_. |
the origin of the ultra - high energy cosmic rays ( uhecrs ) beyond 100 eev has still been an enigma in modern astrophysics .
uhecrs are frequently accompanied by significant fluxes of energetic neutrinos generated in processes associated with hadronic interactions .
although the source direction of cosmic rays with energy above 60 eev can be identified by an analysis of recent auger data ( abraham et al .
2007 ) , neutrinos are regarded as being the most useful probe of their source confirmation owing to no deflection by galactic / intergalactic magnetic fields and the weakest reduction by interactions with the cosmic background radiation .
the models for the origin of the highest energy cosmic rays are classified into two categories : top - down and bottom - up . in top - down scenarios ,
neutrinos are presumed to be decay or annihilation products of cosmological remnants with the grand unified energy scale of @xmath0 gev . within this framework
, numerous models have been proposed so far , such as topological defects , z - bursts , annihilation of dark matter particles and so on ( for a review , see @xcite ) . in an alternative approach , the bottom - up model is based on the idea that the highest energy cosmic rays are generated by astrophysical objects .
the most plausible candidates are gamma ray bursts ( grb : @xcite ) and active galactic nuclei ( agn : @xcite ) .
in addition to these objects , it is pointed out that uhecrs are also produced via stochastic acceleration in the giant lobes of radio galaxy ( cen a : @xcite ) , which accounts for the past year discovery by the hires and auger collaborations .
supernova remnants , x - ray binaries , mini - quasars ( e.g. @xcite , for a review ) , and soft gamma ray repeaters ( sgr : @xcite ) can also be stellar type neutrino sources , any of which could produce observable fluxes of energetic neutrinos .
if protons are accelerated at these astrophysical objects , neutrinos are expected to be produced in collisions with ambient photons or protons .
some of the above - mentioned models predict neutrino flux at the level of a few events per km@xmath1 year .
the upper range of this estimation seems to be within reach of a first - generation neutrino telescope , such as amanda ( antarctic muon and neutrinos detector array ) . among a wide variety of physics topics to be explored with the neutrino telescope ,
the most important goal is to search for the origin of cosmic radiation , especially that originated from agn and/or grb . for this purpose
, amanda has been primarily optimised in the energy range from tev to pev , targeting not only the diffuse , but point source , flux of energetic neutrinos .
according to an analysis of amanda data , we unfortunately have no evidence for point sources so far , ( woschnagg et al . 2005 ; abbasi et al . 2009 ) , except for a temporal coincidence with an orphan flare of the tev blazar , 1es 1959 + 650 ( halzen & hooper 2005 ) .
the icecube , a cubic kilometre - scale neutrino telescope , which is partly operating and still under construction as a successor to amanda at the same site , is designed to detect the fluxes about 10 - 50 events per km@xmath1 year .
similar event rates are predicted by model calculations assuming that agn or grb are the actual source ( halzen 2005 ) and hence the icecube is expected to certainly detect neutrino fluxes from such point sources . in order to observe a _ guaranteed _ source , however , even the km@xmath2-sized detector should safely operate over a period of at least ten years . as for the cosmic accelerators , the most promising mechanism to lead highest energies and power - law energy spectrum
is considered to be of diffusive shock acceleration ( dsa : @xcite ) .
solar energetic particles and galactic cosmic rays are consistently explained by this mechanism applied to shocks at heliosphere and supernova remnants , respectively ( for a review , see @xcite ) ; these acceleration sites have actually been confirmed by some observations ( koyama et al .
polarization data in the range of radio to optical wavelengths imply that _ in situ _ acceleration of electrons ( and possibly ions ) must also be taking place in the knots of agn jets ( honda & honda 2004a ) or in the hot spots of fanaroff - riley type - ii ( fr - ii ) sources ( probably terminal shocks ) ( perley et al . 1984 ; carilli & barthel 1996 ) . at these sites , magnetic fields in the vicinity of a shock front
plays an essential role in particle acceleration .
a detailed configuration of the fields in extragalactic jets has been revealed by polarization measurements using very long - baseline interferometry .
for example , the quite smooth fields predominantly transverse to the jet axis , are typically observed in the core region of bl lac objects ( 1803 + 784 : @xcite ; 0300 + 470 : @xcite ) .
allowing the propagation of a shock wave through the jets , this implies the establishment of a ( quasi- ) perpendicular shock , which is capable of accelerating cosmic ray particles efficiently .
evidence for large - scale toroidal magnetic fields has been also discovered in the galactic center ( gc ) region ( novak et al .
based on the above results , it seems adequate that system of magnetized filaments has been established in blazar jets as well .
we have suggested a theoretical model to account for the generation of such a large - scale toroidal ( transverse ) magnetic field in astrophysical jets ( honda & honda 2002 ) . in this model ,
huge currents launched from a central engine are broken into many filaments whose transverse sizes are self - adjusted for the electromagnetic current filamentation instability ( cfi : @xcite and references therein ) . in the nonlinear stage of the cfi , the magnetized filaments are often regarded as being strong turbulence that can largely deflect charged particles . allowing shock propagation through the jet ,
the particles are expected to be quite efficiently accelerated for the dsa scenario , which appears to be favorably taking place in the agn jets .
indeed , some knots of a radio galaxy jet are associated with the shock fronts ( m87 : @xcite ) and circumstantial evidence for in - situ acceleration of electrons ( meisenheimer et al .
1989 ) have been found .
it was also pointed that a similar pattern of small - scale quasi - static magnetic fields can also be established by some numerical simulations : e.g. , during the collision of electron - positron plasmas existing in snrs , pulsar winds , grbs , relativistic jets , and so on ( kazimura et al .
1998 ; silva et al . 2003 ) . using a three - dimensional relativistic electromagnetic code ,
nishikawa ( 2003 ) show that non - uniform small - scale magnetic field is generated due to the weibel instability at a jet front propagating through an ambient plasma with / without initial fields .
these dynamics might be involved in some of the knots in the fr - i radio jets , which appear to be a shock established when faster material is overtaking with slower one ( m87 : @xcite ) . in the case of blazars , however , the detailed internal structure of their jets has still not been confirmed because of optical thickness with respect to the observer s line of sight .
another approach to shed light on their configuration is provided by the remarkable short variability timescale of a blazar light curve ( e.g. , mrk 421 : reaching a few minutes , @xcite ) .
this is significantly shorter than the light - crossing time at the black hole horizon , which implies the presence of some substructure in the parsec - scale jet . according to our model configuration of the filamentary jet ,
the strength of the local magnetic field is described by a power - law of the filaments transverse sizes .
if charged particles are injected into this system , they are diffusively accelerated by a collisionless shock being scattered by the field fluctuations .
since the efficiency of the acceleration and loss depends upon the spatial size scales , the local maximum energies of accelerated particles are also characterized by their nearest filament sizes .
the spectrum extending to the x - ray region is attributed to the synchrotron radiation of accelerated electrons . in particular
, the correlation between x - ray and gamma - ray light curves of mrk 421 is well reproduced by a model of the hierachical turbulent structure of the jet ( honda 2008 ) .
the most interesting consequence is that the transition of a hierachical turbulent structure seems to be responsible for the observed patterns of energy - dependent light curves , e.g. , soft lag ( takahashi et al .
1996 ; rebillot et al .
2006 ) or hard lag ( fossati et al . 2000 ) and a tight - correlation mode ( sembay et al .
2002 ) . as for mrk 501 ,
which is also firmly established as a tev @xmath3-ray emitter , the x - ray light curve shows a very rapid flare varying over several 100 seconds ( catanese & sambruna 2000 ) .
it is also revealed that the tev flares correlate with x - ray radiation on timescales of hours or less from multiwavelengths observations ( pian et al .
1998 ; catanese et al . 1997 ; krawczynski et al .
2000 ; sambruna et al .
moreover , the largest shift of the peak energy during the peak - luminosity change was observed among all blazars ( kataoka et al .
2001 ) . summarizing these observational results
, we can say that the acceleration of electrons similar to mrk 421 is also taking place in mrk 501 . presuming a finite ion abundance in the jet of mrk 501 ( rawlings & saunders 1991 ) , such a dsa mechanism operates for arbitrary nuclei ( of course including protons ) as well .
indeed , from the x - ray spectrum of ss 433 jet , emission lines of various elements such as ne , mg , si , and fe were observed ( kotani et al .
1996 ) . in the present paper
, we evaluate the achievable maximum energy of protons in the mrk 501 jet to estimate the resultant neutrino fluxes in accordance with our diffusion and acceleration mechanism in a filamentary jet , which we have proposed and developed in a series of papers ( honda & honda 2004b ; 2005 ; 2007 ) . in section 2 , installing our model structure of magnetized filaments to the jet of mrk 501 , we describe the acceleration proccesses of protons using our novel dsa mechanism . taking competitive energy losses and restrictions into consideration , we present the scaling of the maximum energy . in
2.1 we describe the acceleration taking place at local magnetic fields induced by the current filament each ( referred to as preliminary acceleration ) and in 2.2 the additional acceleration due to the inter - filaments deflection . in 2.3 , we calculate the achievable maximum energies of accelerated protons with respect to the transverse filament sizes and magnetic field parameters .
we then evaluate neutrino flux from mrk 501 and discuss its detectability based on a km - scale neutrino telescope such as icecube in section 3 .
conclusions are summarized in section 4 .
suppose a parsec - scale blazar jet transporting energetic particles from the central engine .
the directional plasma flow will favorably induce huge currents ( appl & camenzind 1992 ) that breaks up into many filaments because of electromagnetic cfi .
each filament generates a small - scale transverse magnetic field whose strength is determined by the transverse size of a filament , @xmath4 , i.e. , @xmath5 , where @xmath6 ; @xmath7 and @xmath8 denote the jet diameter and power - law spectral index of the magnetized filamentary turbulence , respectively ( honda & honda 2007 ) . viewing the jet globally from outside ,
randomly oriented fields inside jet are cancelled , except for a large - scale toroidal magnetic field along the envelope .
as was illustrated in our previous paper ( honda & honda 2005 ) , there exists an energy hierarchy for protons trapped in such a filamentary current system : ( i ) @xmath9 and ( ii ) @xmath10 , where @xmath11 and @xmath12 are the charge of an electron and the vector potential , respectively . the former corresponds to a low - energy regime , in which protons are strictly bounded for the local magnetic field induced by each filament .
the latter reflects the higher energy regime in which the validity condition of the quasilinear approximation , @xmath13 is satisfied , where @xmath14 and @xmath15 are the averaged and fluctuated part of the momentum distribution function for protons in the test - particle approximation , @xmath16 , respectively . in this regime , protons are no longer trapped with local magnetic fields , but are deviated with small fluctuations .
below , we describe the acceleration and competitive energy losses in each energy hierarchy . in the low - energy regime of @xmath17 , protons are strictly bounded and gyrating around the local magnetic field induced by each filament .
energies of injected protons are elevated via the conventional diffusive shock acceleration being resonantly scattered from small magnetic fluctuations ( drury 1983 ) .
the characteristic acceleration time of protons is described by @xmath18 where @xmath19 , presuming kolmogorov turbulence , @xmath20 ( @xmath21 ) is the energy density ratio of fluctuating to local mean magnetic fields , @xmath22 is the proton gyroradius , and @xmath23 ( = 4 for the strong shock limit ) is the shock compression ratio .
the achievable maximum energies of protons are limited by the various radiative cooling timescales , and here we consider two representatives : the synchrotron radiation losses , @xmath24 , and collisions with ambient photons , @xmath25 . in blazar jets
the propagation time of shock , @xmath26 , also restricts the highest energy , and hence the temporal limit is given by @xmath27 besides the temporal limit , there exists a spatial limit : the maximum gyroradii of protons are not allowed to exceed the transverse sizes of in - situ filaments . the achievable energy of protons via the preliminary acceleration is determined by comparing the temporal and spatial limit . in the following , conceivable energy constraints for mrk 501 are provided .
gyrating around the filament - induced local fields , and being scattered by small fluctuations , protons emit synchrotron photons , which can be a dominant cooling process when the energy density of the magnetic field is sufficiently greater than that of the radiation .
the characteristic loss time for proton synchrotron is given by ( rybicki & lightman 1979 ) @xmath28 where @xmath29 and @xmath30 are the masses of a proton and of an electron , respectively ; @xmath31 is the cross section for thomson scattering and @xmath32 is the lorentz factor of accelerated protons .
the average energy density of the local magnetic field is denoted by @xmath33 , where @xmath34 .
it should be noted that the adopted field value of @xmath35 g ( tavecchio & maraschi 2001 ) is not the one averaged over the compact blob but the maximum for a filament whose radial size is compared to the jet diameter .
we consider proton - photon collisions leading to a pion - production cascade .
since the collision timescale is characterized by the target photon spectrum , which is still unknown , we adopt here a description by a single power - law : @xmath36 ( @xcite ) , where @xmath37 is the number density of photons per unit energy interval @xmath38 .
then , the timescale can be expressed as @xmath39 where @xmath40 erg@xmath41 ( @xcite ) is the average energy density of target photon fields and @xmath42 ( @xcite ) . since @xmath43 , it is found that @xmath44 .
we should also check whether collisions with particles ( especially , with protons ) becomes effective or not .
the characteristic cooling time of relativistic protons due to inelastic pp collisions can be written as ( @xcite ) @xmath45 where the number density of the target hydrogen medium , @xmath46 @xmath41 , corresponds to the upper limit assuming mass loaded hadronic jet models .
the averaged total cross section , @xmath47 , is approximately 40 mb at very high energies , and @xmath48 is the coefficient of inelasticity
. it would be fair to say that collisions with protons should not be taken into consideration .
the achievable maximum energy of protons is also restricted by the duration that the shock is propagating through the jet from the central engine to the working surface .
this corresponds to the age of a blob , which is propagating with mild - relativistic speed through the relativistic jet ( @xcite ) , and is approximately written by @xmath49 . for mrk 501
, we obtain @xmath50 where @xmath51 pc is the distance from the core to the blob ( @xcite ) and @xmath52 is the average speed of the blob ( @xcite ) .
this blob is currently operating for particle acceleration , and therefore @xmath26 can not be compared to the adiabatic loss timescale for agn , which corresponds to the lifetime of the shock ( @xcite ) .
the timescale of adiabatic losses is expressed as @xmath53 , where @xmath54 and @xmath55 are the lorentz factor of the jet and the speed of the radial expansion .
considering narrow opening angles of blazars , which implies @xmath56 , we can safely say that @xmath26 is sufficiently shorter than @xmath57 .
once accelerated particles acquire sufficient energies , they are escaping from the local magnetic fields , given as a function of the filament sizes .
this spatial limit is given by the assumption that the gyroradius of each particle should not be beyond the transverse size of the local filament .
then , the local maximum energy is denoted by @xmath58 .
the transverse sizes of the minimum and maximum filaments are regarded as being comparable to the debye sheath and the jet diameter , respectively .
thus , the energy of particles escaping from the largest filament can be written as @xmath59 where @xmath60 pc is the extent of the radio emitting region ( @xcite ) .
this value is estimated by substituting @xmath61 g for @xmath62 in the formula suggested by marscher ( 1987 ) .
once protons are accelerated and injected into the high - energy regime of @xmath63 , they are no longer bounded to the local magnetic field .
field vectors are randomly oriented in the internal jet in the transverse direction to the filament , and hence these protons can move almost freely in the jet being deflected by the random magnetic field . allowing for shock propagation , protons are additionally accelerated in this regime , being off - resonantly scattered in a forest of magnetized filaments . in this aspect , the preliminary accelerator via the conventional dsa mechanism can be regarded as being an injector to this further booster .
the new injection mechanism is referred to as transition injection , in analogy to bound - free transition in atomic excitation .
the characteristic time of the additional acceleration for protons is given by @xmath64 where @xmath65 ( e.g. , @xcite ) and @xmath66 are the spectral index and energy of protons , respectively ; @xmath67 is the filamentary turbulent spectral index , and @xmath68 is assumed . the maximum energy boosted by the additional acceleration is also restricted by some radiative loss processes .
in addition to two possible timescales : for diffuse synchrotron , @xmath69 , and for the p@xmath3 collision , @xmath70 , we consider the shock propagation timescale , @xmath26 , and therefore @xmath71 .
the spatial limit will also be considered . the additionally accelerated protons deflected by the random magnetic fields emit unpolarized synchrotron radiation . here
, we adopt the characteristic cooling time for the diffuse synchrotron from our previous paper ( @xcite ) , which was derived from the theoretical basis by @xcite : @xmath72 where @xmath73 is the timescale of normal synchrotron losses and @xmath74 $ ] . for @xmath8 ( @xcite ) , the maximum energy of a proton , @xmath32 ,
is directly estimated by the equation @xmath75 , which gives @xmath76 it should be noted that the maximum energy restricted by the diffuse synchrotron is free from the magnetic field strength because of the same @xmath62-dependence of @xmath77 and @xmath69 .
cooling due to collisions with photons is also characterized by the above - mentioned diffuse synchrotron : @xmath78 . using the same target photon spectrum and other notation with those given in @xmath79
, we obtain @xmath80 as we have mentioned in 2.1.3 , the shock propagation time regarded as the age of the blob restricts the acceleration .
equating @xmath81 , we obtain @xmath82 it should be noted that additionally accelerated particles are not gyrating to the local fields , but deflected by randomly oriented fields .
thus , the spatial limit is determined by the condition that the mean free path for this collisionless deflection should be smaller than the system size : @xmath83 , where @xmath84 is a dimensionless parameter ( see , @xcite , for the details ) .
it gives @xmath85 where @xmath86 is assumed .
putting together the above discussion , provided both the normal and transition injection of protons are realized in a blazar jet , multi - stage diffusive shock acceleration can take place . in this subsection , we present the results of a numerical evaluation of the achievable maximum energies of protons in both stages within the conceivable parameter region . in fig.[fig :
fig1 ] , the energies of accelerated protons are plotted against the transverse filament size parametrizing the ratio of the mean / fluctuate magnetic energy density , @xmath20 .
the transverse filament sizes , @xmath4 , covers the range from @xmath87 pc to @xmath88 pc .
the minimum and the maximum sizes are compared to the correlation length of alfvenic fluctuation and the jet diameter , respectively . as shown , in the case of preliminary acceleration the achievable maximum energies of protons
are mainly restricted by the escape from individual filaments , which is represented by a linearly increasing line ( @xmath89 ) . in the case of @xmath90 ,
the proton energies achieved via preliminary acceleration , @xmath91 , are determined by @xmath92 over the whole range of the considered filament sizes . in the case of @xmath93 , the values of @xmath91
are also determined by @xmath92 for @xmath94 pc , while the @xmath91-values are more severely restricted by the propagation time of the shock for @xmath95 pc , and hence the achievable energy via the preliminary acceleration , @xmath91 , is cut - off at @xmath96 ev . if @xmath97 , the @xmath91-values are perfectly overlapped with those of @xmath90 .
on the other hand , the maximum energy achieved by the additional acceleration @xmath98 is independent of not only the filament size , but also the @xmath20-parameter .
the energies of accelerated protons can be additionally boosted up to the value determined by the spatial limit expressed with a horizontal line of @xmath99 ev , irrespective of their energies acquired by the preliminary acceleration .
( 80mm,80mm)figure1.eps we also present the @xmath4-dependence of @xmath91 parametrizing the @xmath100-value in fig.[fig : fig2 ] .
the lines corresponding to @xmath101 shift upwards with a larger @xmath100 .
for the cases of @xmath102 g and 0.1 g , the proton energies achieved by the preliminary acceleration , @xmath91 , are determined by @xmath92 over the whole range of conceivable filament sizes . in the cases of @xmath103 g and 10 g ,
the protons dominantly lose their energies via collisions with photons for @xmath104 pc and for @xmath105 pc , respectively .
thus , the resultant maximum value of @xmath91 is nearly @xmath106 ev . as for the additional acceleration ,
the @xmath98-values are expressed by horizontal lines because of their independence of @xmath4 .
the lowest and the highest @xmath98 correspond to those of @xmath107 g and @xmath108 g , respectively , and in both cases the maximum energies are restricted by the spatial limit .
the @xmath98-value for @xmath103 g is identical to that for @xmath109 g , and in these cases the energy is restricted by collisions with ambient photons .
in the previous section we calculated the achievable maximum energies of protons in a blazar jet . since the proton energies accelerated by multi - stage dsa are sufficiently higher than the neutrino - producing threshold via photopionization , we can safely expect significant fluxes of energetic neutrinos . in the photomeson process of @xmath110
, the center - of - mass energies of accelerated protons should be beyond the @xmath111-resonance threshold , @xmath112 mev , @xmath113 where @xmath54 and @xmath114 are the lorentz factor of a blazar jet and the mean energy of ambient photons , respectively .
presuming that the generated pion energy is equally divided into four leptons through the decay @xmath115 , the neutrino energy is described as @xmath116 where @xmath117 is the averaged fraction of momentum transfer from a proton to a pion .
as for the mrk 501 , the peak energy of synchrotron emssion in the x - ray band is shifted from 1 kev to 100 kev during flaring ( @xcite ) .
since these energies are achievable for synchrotron photons emitted from co - accelerated electrons , we take these values as the mean energy of ambient photons . for @xmath118 kev
, we obtain the neutrino producing threshold using eq.([eqn : epmin ] ) as @xmath119 which leads to @xmath120 tev .
similarly , for @xmath121 kev , @xmath122 pev and hence @xmath123 pev . since the maximum energy of accelerated protons is sufficiently higher than these threshold values , mrk 501 can be a feasible source of neutrinos . in both cases we determined @xmath124 pev , according to the upper limit estimated in 2.3 . in order to calculate the neutrino flux produced at an individual blazar
, we begin with the following useful formula : @xmath125 where the upper and lower limits of integration are the maximum and minimum ( threshold ) neutrino energies , estimated in 3.1 , respectively .
the luminosity distance to the blazar , @xmath126 , is defined as @xmath127 where @xmath128 is the redshift ( = 0.034 for mrk 501 : @xcite ) for mrk 501 and @xmath129 is the proper motion distance , which is written by the hubble constant , @xmath130 km@xmath131mpc@xmath131 , the matter density , @xmath132 , and dark energy density , @xmath133 , adopted from wmap results ( @xcite ) .
the observed neutrino luminosity , @xmath134 , is defined as @xmath135 where @xmath136 , @xmath137 , and @xmath138 are the number of produced neutrinos , the mean neutrino energy , and the observed variability time of the emission region , respectively .
since @xmath136 and @xmath137 are not observables , we introduce the optical depth defined as @xmath139 where @xmath140 erg / s is the observed photon luminosity , @xmath141 represents the volume of emission region with radius @xmath142 , @xmath143 ev denotes the mean energy of photons , and @xmath144 cm@xmath1 is the @xmath111-resonance photopionization cross section .
then , we can express @xmath134 with observable quantities in a compact notation , @xmath146 where @xmath147 is a constant taking the branching ratios of interaction chain and corresponding momentum transfer into account .
the intrinsic proton luminosity , @xmath148 , can be replaced with the observed @xmath149 taking the proton interactions with ambient photons into consideration .
however , @xmath149 is still unknown ; we assume it to be as 10 @xmath150 of the photon total luminosity ( @xcite ) .
we also replace @xmath151 with @xmath152 , with considering the possible p and @xmath153 absorption in the vicinity of the source . summarizing the above discussion , we adopt the following formula of differential flux from an individual blazar ( @xcite ) : @xmath154 where the factor 2 reflects the contribution of neutrinos and anti - neutrinos .
since we are concerned with the point - source neutrino flux , we consider the @xmath155 contribution on account of their highest angular resolution .
this is attributed to the large muon track , having advantege of reconstructing paths . in fig.[fig
: fig3 ] we plot the differential flux of neutrinos from mrk 501 against their energies .
the energy of neutrinos is limited to the range between the maximum and the minimum predicted by this model . here
, we adopt the typical power - law energy distribution of neutrinos , @xmath156 , with the exponential cutoff , @xmath157 .
we parametrize the power - law index @xmath158 , 2.0 , 2.7 , where @xmath158 and @xmath159 are taken from the upper and lower limits of the exponents , respectively , and @xmath160 is a resemblance of the parent proton spectrum supposed to be accelerated by fermi s mechanism . the cutoff energy , @xmath161 gev , is the achievable maximum energy of agn protons estimated in @xmath162 .
one can see that for a steeper energy spectrum the neutrino flux is more significantly reduced in the energy range of @xmath163 gev . on the contrary , in the lower energy region of @xmath164
gev the flux is higher in the case of a steeper spectrum because of the contribution from the denominator .
anyhow , the neutrino flux originating from a single blazar is trivial compared to the observable level .
( 80mm,80mm)figure3.eps it is more difficult to detect higher energy neutrinos because of the steeply falling spectrum . in order to detect the interaction of a tev neutrino with a cherenkov telescope , the effective volume of the detector is required for kilometer scale to cover the typical range of tev muons . in the effective region of the telescope ,
the probability to detect @xmath155 in the tev - pev range is approximately given by @xmath165 where @xmath166 and @xmath167 are the muon range and the neutrino interaction length , respectively ( @xcite ) .
we can now compute the diffuse neutrino event rate by integrating the total differential flux multiplied by the detection probability of eq.([eqn : prob ] ) @xmath168 where the upper and lower limits of the integral are @xmath169 ev and @xmath170 ev calculated in @xmath171 3.1 , respectively . in order to compute the diffuse neutrino flux from the observed blazar distribution
, we should integrate the flux from all blazars while taking the doppler factor distribution into consideration . for simplicity
, here we adopt the effective number of blazars ( @xcite ) , using the @xmath3-ray flux ratio of diffusive to the single blazar , which is derived from the luminosity function of 20 brightest blazars obtained by the egret collaboration ( @xcite ) .
thus , instead of summing up the total isotropic differential flux of neutrinos @xmath172 , we simply estimate the diffuse flux by multiplying the flux of mrk 501 by the resultant number 130 sr@xmath131 . a correction for the difference in the spectral indices of gamma - ray and neutrino fluxes enhances the neutrino flux by a factor of three .
we present the estimated neutrino event rates in table 1 for conceivable power - law exponents . considering the 4@xmath173 coverage and typical exposure time ( @xmath174 yr ) , the value @xmath175
km@xmath176 yr@xmath131 sufficiently reaches to the observable level .
.neutrino event rates for various power - law exponents.[tbl-1 ] [ cols="^,^",options="header " , ]
in this paper we applied our model of multi - stage diffusive shock acceleration to the blazar mrk 501 and evaluated the maximum energies of protons .
we also calculated the neutrino flux from this single source generated by these accelerated protons , and discuss the detectability with the updated neutrino telescope , icecube .
we obtained the eev protons and differential flux during the operating time of the detector by not a simple , optimistic order estimation , but by considering the systematical energy restriction at the source .
the results are rather sensitive to the values of the magnetic field strength and the size of the accelerator .
the @xmath62-field value is related to the system size via the radiation intensity of the emission region . in the conceivable range of the field strength ,
eev protons are obtained by additional acceleration .
giovannini , g. , feretti , l. , venturi , t. , cotton , w. d. , & lara , l. 1999 , in asp conf .
l. o. takalo & a. s. silanp ( san fransisco asp ) , bl lac phenomenon , a conference held 22 - 26 june 1998 in turku , finland , 439 tavecchio , f. , & maraschi , l. 2001 , in aip conf .
proc . , 599 , x - ray astronomy : stellar endpoints , agn , and the diffuse x - ray background , ed . n. e. white , g. malaguti , & g. g. c. palumbo , ( new york : aip ) , 979 | blazars have been regarded as one of the most powerful sources of the highest energy cosmic rays and also their byproducts , neutrinos . provided that a magnetized filamentary system is established in a blazar jet as well
, we could apply the mechanism of multi - stage diffusive shock acceleration to a feasible tev emitter , mrk 501 to evaluate the achievable maximum energy of protons . taking conceivable energy restriction into account systematically
, it seems adequate to say that eev - protons are produced at this site by our present model .
we also estimate neutrino fluxes generated by these accelerated protons and discuss the detectability based on an updated kilometre - scale telescope such as icecube . |
nonadiabatic transition due to potential curve crossing is one of the most important mechanisms to effectively induce electronic transitions @xcite .
this is a very interdisciplinary concept and appears in almost all fields of physics and chemistry , even in biology @xcite .
the common examples are variety of atomic and molecular processes such as atomic and molecular collisions , chemical reactions and molecular spectroscopic processes .
attempts have been made to classify organic reactions by curve crossing diagrams @xcite .
there are also examples in nuclear physics @xcite .
two state curve crossing can be classified into the following two cases according to the crossing scheme : ( 1 ) landau - zener ( l.z ) case , in which the two diabatic potential curves has the same signs for the slopes and ( 2 ) non - adiabatic tunnelling ( n.t ) case , in which the diabatic curves has the opposite sign of slopes . in our earlier paper @xcite paper
we have considered two diabatic curves , crossing each other and there is a coupling between the two curves , which causes transitions from one curve to another .
this transition would occur in the vicinity of the crossing point .
in particular , it will occur in a narrow range of x , given by @xmath0 where @xmath1 denotes the nuclear coordinate and @xmath2 is the crossing point . @xmath3 and @xmath4 are determined by the shape of the diabatic curves and @xmath5 represent the coupling between them .
therefore it is interesting to analyse a model , where coupling is localized in space near @xmath2 .
thus we put @xmath6 here @xmath7 is a constant .
this model has the advantage that it can be exactly solved @xcite . in this paper
we apply this model in the case of general multi - channel problems ( fig .
[ mcurve ] ) , since this is very important in view of the applications to various practical problems in chemical dynamics beyond simple diatomic systems .
using model potential in which one linear diabatic potential intersects many constant diabatic potentials , demkov and osherov @xcite showed that the state to state transition probabilities can be exactly calculated by a product of the landau - zener probabilities at each crossing .
a more general model , in which one set of parallel linear diabatic potentials intersect with another set of parallel ones , has been analyzed by child @xcite .
korsch @xcite reported multi - state curve crossing problems by the magnus approximation , and later nakamura @xcite analyzed the problems by complex phase - integral technique . in the following we give a general procedure for finding the exact quantum mechanical solution for multi - channel curve crossing problem by extending our two state model @xcite
the time dependent schr@xmath8 dinger equation ( we take @xmath9 here and in subsequent calculations ) @xmath10 @xmath11 is defined by @xmath12 where @xmath13 is @xmath14 and @xmath15 @xmath16 and @xmath17 are the probability amplitude for the two states .
we find it convenient to define the half fourier transform @xmath18 of @xmath19 by @xmath20 half fourier transformation of eq .
( [ 4 ] ) leads to @xmath21 this may be written as @xmath22 @xmath23 is defined by @xmath24 @xmath25 . in the position representation , the above equation may be written as @xmath26 where @xmath27 is @xmath28 writing @xmath29 and we can write @xmath30^{-1}|x_0\rangle.\ ] ] this expression simplify considerably if @xmath31 is a delta function located at @xmath2 . in that case
@xmath31 may be written as @xmath32 @xmath33 @xmath34 @xmath35 .
then @xmath36^{-1}|x_0\rangle , \ ] ] where @xmath37 now we use the operator identity @xmath38^{-1}.\ ] ] inserting the resolution of identity @xmath39 @xmath40 @xmath41 in the second term of the above equation , we arrive at @xmath42 putting @xmath43 in eq . (
[ 20 ] ) and solving for @xmath44 gives @xmath45^{-1}.\ ] ] this when substituted back into eq .
( [ 20 ] ) gives @xmath46 using the same procedure one can get @xmath47 similarly one can derive expressions for @xmath48 and @xmath49 .
we start with the time dependent schr@xmath8dinger equation for a three state system ( we put @xmath9 ) @xmath50 the above equation can be splitted into @xmath51 and @xmath52 and also @xmath53 we define half fourier transform @xmath54 and fourier transform of the above three equations gives @xmath55 and @xmath56 as @xmath57 .
also @xmath58 as @xmath59 we will substitute @xmath60 and @xmath61 in eq . ( [ 55a ] ) @xmath62^{-1}v_{21}(x)\psi _ 1(x)\ ] ] and @xmath63^{-1}v_{31}(x)\psi _ 1(x)\ ] ] eq .
( [ 55a ] ) becomes @xmath64^{-1}v_{21}(x)-v_{13}(x)[\omega -h_{33}(x)]^{-1}v_{31}(x)]% \overline{\psi } _
1(x,\omega ) = i\psi _ 1(x,0)\ ] ] the above equation simplify considerably if the couplings are assumed to be dirac delta function , which in operator notation may be written as @xmath65 and @xmath66 .
the above equation may be written as @xmath67\overline{\psi } _
1(x,\omega ) = i\psi _ 1(x,0)\ ] ] or @xmath68\overline{\psi } _
1(x,\omega ) = i\psi _ 1(x,0)\ ] ] where , @xmath69 the solution of this equation may be expressed in terms of the green s function @xmath70 @xmath71g^{\prime } ( x , x_0;\omega ) = i\delta ( x - x_0)\ ] ] also @xmath72 here @xmath70 is @xmath73 where @xmath74 in the previous section we have calculated @xmath75 , which is redefined below @xmath76 the @xmath75 can be written @xmath77 in this case , one can find the green s function @xmath75 , using the method as we have used before .
@xmath78 so in this way one can add the effect of 4-th state , 5-th state etc .
one by one and calculate green s function for @xmath79-channel problem . here
we start with @xmath79-channel generalization of eq .
( [ 55a ] , [ 55b ] , [ 55c ] ) @xmath80 and @xmath81 from the eq .
( [ 74 ] ) we calculate @xmath82 @xmath83^{-1}v_{\varepsilon 1}(x)\overline{\psi } _
1(x,\omega ) \ ] ] so eq . (
[ 73 ] ) can be written as @xmath84^{-1}v_{\varepsilon 1}(x)% \overline{\psi } _
1(x,\omega ) \ ] ] we take @xmath85 .
the above expression becomes @xmath86 where @xmath87^{-1}|x_\varepsilon \rangle \\
= \sum\limits_{n=0}^\infty \langle x_\varepsilon |\phi _ n\rangle [ \omega -h_{\varepsilon \varepsilon } ( x)]^{-1}\langle \phi _ n|x_\varepsilon \rangle \end{array } \\
= \sum\limits_{n=0}^\infty \phi _ n^\varepsilon ( x_\varepsilon ) [ \omega -h_{\varepsilon \varepsilon } ( x)]^{-1}\phi _ n^{\varepsilon * } ( x_\varepsilon ) \\ \end{array}\ ] ] in the above expression , @xmath88 takes discrete values .
now @xmath89\overline{\psi } _
1(x,\omega ) = i\psi _ 1(x,0)\ ] ] the solution of this equation may be expressed in terms of the green s function @xmath90 @xmath91g^{\prime \prime } ( x , x_0;\omega ) = i\delta ( x - x_0)\ ] ] here @xmath90 is @xmath92^{-1}|x_0\rangle .$ ] also @xmath93 in the continuum limit ( here @xmath88 changes continuously ) , the expression for @xmath90 can be written @xmath94g^0\{x(\varepsilon ) , x(\varepsilon ) ; \omega , \varepsilon \}]^{-1}|x_0\rangle\ ] ] in a very special case where all states couples at one point to the first one i.e. @xmath95 @xmath96^{-1}|x_0\rangle\ ] ] the above expression is very general and applicable to any potential .
we can write eq .
( [ 80b ] ) in a simplified form @xmath97^{-1}|x_0\rangle\ ] ] where @xmath98\ ] ] : in the following , we discuss the case where one potential ( general ) is coupled to continuum ( in energy ) of constant potentials at a point .
green s function for uncoupled potential is given by @xmath99 in the above expression @xmath88 varies continuously from @xmath100 to @xmath101 .
analytical form of @xmath102 depends on the functional form of @xmath103 . for our calculation
we take @xmath104 so @xmath105 the eq .
( [ 81 ] ) now becomes @xmath106^{-1}|x_0\rangle\ ] ] the above expression can be evaluated exactly .
we have extended our two state model to deal with general one dimensional multi - channel curve crossing problems in presence of delta function couplings .
we have analyzed the case where a potential energy curve couples to a continuum ( in energy ) of potential energy curves .
nonadiabatic tunneling in an ideal one dimensional semi - infinite periodic potential system is also analyzed using our model .
the same procedure is also applicable to the case where @xmath107 is a non - local operator , and may be represented by @xmath108 @xmath7 @xmath109 , @xmath110 and @xmath111 are arbitrary acceptable functions . choosing both of them to be gaussian will be an improvement over the delta function coupling model .
@xmath107 can also be a linear combination of such operators .
the author thanks prof .
k. l. sebastian for suggesting intersting problems .
h. nakamura , int .
phys . chem .
* 10*,_123 _ ( 1991 ) .
h. nakamura , in _ theory , advances in chemical physics _ , edited by m. bayer and c. y. ng ( john wiley and sons , new york , 1992 ) .
s. s. shaik and p. c. hiberty , edited by z. b. maksic , _ theoretical models of chemical bonding , part 4_,(springer - verlag , berlin , 1991 ) , vol . 82
. b. imanishi and w. von oertzen , phys . rep . * 155 * , _
29_(1987 ) .
a. thiel , j. phys .
g * 16 * , _ 867 _ ( 1990 ) .
a. chakraborty , mol .
( 2009 ) ( to appear ) .
n. demkov and v. i. osherov , soviet phys .
jetp * 26 * , _ 916 _ ( 1968 ) .
m. s. child , faraday discussions chem soc .
, * 55*,_30 _ ( 1973 ) . h. korsch , j. mol . phys .
* 201*,_51 _ ( 1980 ) . | we have proposed a general method for finding the exact analytical solution for the multi - channel curve crossing problem in presence of delta function couplings .
we have analyzed the case where a potential energy curve couples to a continuum ( in energy ) of potential energy curves . |
in the context of molecular simulations , hybrid quantum mechanics - molecular mechanics ( qm - mm ) schemes consider the system as a sum of two parts : solute ( qm fragment ) and solvent ( mm fragment ) @xcite .
the particles are assigned to one of these two groups according to their role : atoms directly involved in bonds breaking or forming , or in polarization or charge transfer effects , must be considered in the qm region , whereas those atoms not participating in these processes are included within the mm subsystem .
these two groups are described at different levels with different hamiltonians , but they interact with each other , generally self - consistently .
qm - mm schemes have been applied extensively and successfuly along the last couple of decades , to model finite chemical and biological systems .
the impact of this methodology has been acknowledged through the chemistry nobel prize of the year 2013 , which was awarded to some of its founders for the development of multiscale modelling .
one of the major successes of this approach was in the study of chemical reactions inside the active site of proteins . in this type of simulations
the solute described quantum - mechanically comprises the active site , while the rest of the protein plus hydration water molecules are treated classically @xcite .
leaving aside the applications in biochemistry , hybrid qm - mm methodologies in a non - periodic setting have also been employed in various other contexts , as for example proton transfer reactions in water clusters @xcite , or in different kinds of materials which were modeled as finite structures : these works have addressed solid - liquid @xcite , metal - organic @xcite , and oxide interfaces @xcite . on the other hand ,
the qm - mm methodology applied in periodic boundary conditions ( pbc ) for both the qm and mm parts , has been rarely reported in the literature . a few examples imposing periodicity to the mm region only to model
dilute solutions can be found for semiempirical or first - principles approaches @xcite .
laino et al .
@xcite developed a qm - mm method in periodic boundary conditions based on gaussian basis sets and multigrids to treat the long - range interactions , which was tested on the simulation of surface defects present at the @xmath1-quartz phase of silica .
other periodic qm - mm implementations with gaussian basis sets have been proposed based on the reduction of the electron density to point charges , after which the classical ewald summation can be applied @xcite .
such strategy has been implemented for both semiempirical @xcite and ab - initio @xcite hamiltonians . in this line ,
golze and co - workers elaborated a method for the treatment of metallic interfaces , where the interactions between the quantum - mechanical adsorbate and the classical substrate are handled at the molecular mechanics level @xcite .
to the best of our knowledge only yarne et al .
@xcite developed a hybrid qm - mm methodology imposing pbc to the whole system in a pseudopotentials planewaves ( ppw ) code @xcite . in this case electrons
were confined to a smaller unit cell inside the supercell needed to describe the whole system , and periodicity was limited to 1 or 2-d @xcite . in the present article , we present a formulation for hybrid qm - mm calculations based on density functional theory ( dft ) in a ppw framework .
in particular , this scheme has been devised for the car - parrinello method as implemented in the quantum espresso code @xcite .
the goal is to have available a hybrid qm - mm methodology in pbc appropriate to describe condensed matter in complex environments , as solid - liquid or liquid - liquid interfaces , where the main interest or the `` chemistry '' involves one of the two phases the qm part under the influence of the other the mm part
. this could be useful in solid - water interfaces , where the solid and any adsorbed species can be described quantum - mechanically , while the solution may be modeled using a classical force - field .
we illustrate this kind of application through the study of titania in contact with an aqueous phase .
the opposite representation , in which the solid constitutes the mm part , might also be appealing if the interest were in the properties of the other phase , as it could be the case of nanoconfined molecules or fluids .
to the best of our knowledge , no other qm - mm model has been based on the present strategy , which we believe is a very interesting one for atomistic simulations of interfaces or nano - spaces with high accuracy at an affordable computational cost .
in the dft - ppw approach used in the quantum espresso code , the qm energy can be cast as @xcite : @xmath2 + e_h[\rho ] + e_{ii } + e^{loc}_{ps}[\rho ] + e_{ps}^{nl } + e_{xc}[\rho ] \label{eqmtot}\ ] ] on the right hand side of the above equation , from left to right , there is the kinetic energy of the electrons , the hartree energy , the ion - ion repulsion , the local and non - local contributions to the pseudopotential energies , and the exchange - correlation functional .
here @xmath3 is the electron charge density . in the context of qm - mm models , the hamiltonian and the energy of the system
are written as : @xmath4 @xmath5 where h@xmath6 ( and the related energy e@xmath6 ) is a coupling term describing the interaction between the two regions of the system . in the mm region , atoms are typically treated as point charges of charge @xmath7 interacting with each other through electrostatics , dispersive - repulsive and harmonic potentials , so that the molecular mechanics energy @xmath8 is the sum of three contributions @xcite : @xmath9 where @xmath10 , @xmath11 and @xmath12 denote the electrostatic , the lennard - jones , and the bonding energy respectively , the later of which models the intramolecular degrees of freedom . in turn , these terms are normally computed as : @xmath13 @xmath14 \label{lennardjones}\ ] ] @xmath15 in the second of these three equations @xmath16 and @xmath17 are the lennard - jones radius and interaction energy between atoms @xmath18 and @xmath19 . in the last expression , @xmath20 , and @xmath21 ,
represent force constants for the harmonic potentials controlling bond lengths , angles and torsions , respectively . within the mm region
we will consider only water molecules , which internal degress of freedom are described through the o - h distances and h - o - h angles , and therefore the third term in the last equation will not be present .
the @xmath22 contribution appearing in equation [ energy - qmmm1 ] can normally be explicitely written as the sum of an electrostatic and a non - electrostatic term . as it will be shown below , however , in the working formula implemented here the electrostatic contribution to @xmath22
can not be written separately , because it is intertwined with the total electrostatic energy . in the ppw method ,
the electrostatic contribution comes from the sum of the second , third , and fourth terms on the right hand side of equation [ eqmtot ] : @xmath23 = e_h[\rho ] + e^{loc}_{ps}[\rho ] + e_{ii}\ ] ] @xmath24 where @xmath25 indicates the atomic species , @xmath26 is the number of different atomic species , and @xmath27 is the local part of the pseudopotential for each species .
@xmath7 is the ionic charge of the nuclei ( which amounts to the atomic number minus the valence electrons ) and * * r**@xmath28 their positions .
@xmath29 stands for the number of ions and @xmath30 for the number of ions corresponding to the atomic species @xmath25 .
we adopt the same convention used in the computational code , in which the sign of the electronic charge is taken as positive and the ionic charge as negative . due to the long - range decay of electrostatic interactions , @xmath31 $ ] ,
@xmath32 $ ] and @xmath33 diverge if they are calculated separately .
it turns out to be convenient to introduce a fictitious ionic charge , @xmath34 , which is added to the hartree energy term and substructed from the other two . in this way ,
a total neutral charge density is defined , @xmath35 , and the electrostatic energy can be rewritten as : @xmath23 = \dfrac{1}{2 } \iint { \dfrac{\rho_t(\textbf{r})\rho_t(\textbf{r'})}{|\textbf{r}-\textbf{r'}| } d\textbf{r}d\textbf{r ' } } + \int { \rho(\textbf{r } ) \left ( \sum_{s=1}^{n } { \sum_{l=1}^{p_s } { v_{ps}^{loc , s } ( |\textbf{r } - \textbf{r}_i| ) d\textbf{r } } } - \int { \dfrac{\rho_{\alpha}(\textbf{r'})}{|\textbf{r}-\textbf{r'}| } } \right ) d\textbf{r } } \ ] ] @xmath36 the hartree and the local pseudopotential contributions can be expanded in fourier space , and the ion - repulsions treated with the ewald method . in particular , if the ionic charge is defined as a sum of gaussian functions centered on every nuclei , @xmath37 after some manipulation the electrostatic energy can be expressed as @xcite : @xmath23 = \dfrac{\omega}{2 } \sum_{\textbf{g } } \dfrac{4\pi}{g^2 } \tilde{\rho}_t ( \textbf{g } ) \tilde{\rho}_t ( -\textbf{g } ) + \omega \sum_{\textbf{g } } \left [ \sum_{s=1}^{n } s_s(\textbf{g } ) \tilde{v}_{ps}^{loc , s}(\textbf{g } ) -\dfrac{4\pi}{g^2}\tilde{\rho}_\alpha(\textbf{g } ) \right ] \tilde{\rho } ( -\textbf{g})\ ] ] @xmath38 - \dfrac{\eta}{\sqrt{\pi } } \sum_{i=1}^{p } z^2_i .\ ] ] where @xmath39 is the volume of the supercell , @xmath40 is a cutoff distance parameter , and @xmath41 is an atomic structure factor for each species @xmath25 , @xmath42 in the last couple of equations , @xmath43 , and @xmath44 are the coefficients of the fourier expansions of the corresponding real space functions , with @xmath45 the reciprocal lattice vectors ( @xmath46 ) .
the complementary error function @xmath47 arises from the point - charges interactions screened by the gaussian functions , with @xmath48 an index running over cells in real space .
the short - ranged nature of this interaction ensures that the sum converges very fast : typically , only first neighbours need to be considered .
one of the key points in our hybrid approach is to conceive the mm atoms in the same way as the pseudoions of the qm region within the ppw framework .
there are basically two differences between mm and qm ions in this case : ( i ) the mm atoms do not include a non - local pseudopotential term , and ( ii ) the mm ions can have a partial charge , which can be either negative or positive , according to the charge parameter in the force field .
hence , in our implementation , the electrostatic energy is extended to include the mm atoms : @xmath23 = e_h[\rho ] + e^{loc}_{ps}[\rho ] + e_{em}[\rho ] + e_{im } \label{qmmm-1}\ ] ] where @xmath49 $ ] and @xmath50 represent , respectively , the interaction of the electron density with the classical charges , and the coulomb interaction between all ions , both qm and mm .
@xmath51 & = \sum_{m=1}^{m } \sum_{l=1}^{p_m } \int { \rho(\textbf{r } ) { { v_{mm}^{m } ( |\textbf{r } - \textbf{r}_i| ) d\textbf{r } } } } \\ e_{im } & = \dfrac{1}{2 } \sum_{i=1}^{t}{\sum_{j=1,j \neq i}^{t}{\dfrac{z_iz_j}{|\textbf{r}_i-\textbf{r}_j| } } } \end{aligned}\ ] ] here @xmath52 and @xmath53 are , respectively , the number of classical species and the number of atoms for the @xmath54 species . the function @xmath55 is the pseudopotential associated with the classical species @xmath54 , to be defined below . *
* r**@xmath28 is the position of every atom , irrespective of being quantum or classical , and @xmath7 is its charge , that will be typically a non - integer number in the mm region .
@xmath56 denotes the total number of atoms in the system ( @xmath57 ) .
the pseudopotential associated with the classical atoms , @xmath55 , has to verify a few properties : has to be a smooth continuous function to be numerically tractable with fast fourier transforms , has to decay as the inverse of the distance @xmath58 at long ranges , and must avoid the divergence when @xmath59 .
we have adopted the functional form proposed by laio et al .
@xcite : @xmath60 with @xmath54 the classical atom species , @xmath61 its charge , and @xmath62 a cutoff radius appropriate for every species .
this function approaches @xmath63 for @xmath64 , and goes smoothly to @xmath65 for @xmath66 .
even if the exact value of @xmath67 at short ranges is not critical , it has to be small enough not to become a trap for the electrons . in the case of plane - wave basis , sharp mm potentials of positive species may cause electronic charge localization on the classical atoms : this is called the _ spill out _ effect .
the possibility of electron density flowing to the mm region can be minimized using a classical pseudopotential which varies softly and has a small magnitude at short distances .
the function defined in equation [ eq - vmm ] satisfies these conditions , providing at the same time an appropriate interaction between mm and qm atoms . with these modifications ,
the electrostatic energy amounts to the following final form : @xmath23 = \dfrac{\omega}{2 } \sum_{\textbf{g } } \dfrac{4\pi}{g^2 } \tilde{\rho}_t ( \textbf{g } ) \tilde{\rho}_t ( -\textbf{g})\ ] ] @xmath68 \tilde{\rho } ( -\textbf{g})\ ] ] @xmath69 - \dfrac{\eta}{\sqrt{\pi } } \sum_{i=1}^{t } z^2_i .\ ] ] this expression is identical to equation [ elec - final ] , aside from the term involving the structure factor @xmath70 corresponding to the mm species , and from the fact that the sums in the last two terms run over mm and qm atoms .
the charge @xmath71 now includes the contribution of the mm ions , and therefore it can take either negative or positive values across space . there is a non - electrostatic contribution to the qm - mm energy which is analogous to that between atoms in the mm region : @xmath72 \label{lennardqmmm}\ ] ] where now @xmath73 and @xmath74 are the parameters for the lennard - jones interaction of a classical atom of species @xmath25 with a quantum atom @xmath54 . this energy prevents the mm charges of negative sign from collapsing on the positive qm nuclei . the mm subsystem in the present study involved h@xmath0o molecules , which were described through the spc flexible water model ( spc / fw ) proposed by wu , tepper and voth @xcite .
the same set of parameters for @xmath75 and @xmath76 were used in both the mm - mm and the qm - mm non - electrostatic interactions , given respectively in equations [ lennardjones ] and [ lennardqmmm ] .
at this point it must be noticed that most force - fields , including the spc / fw potential , do not consider any coulomb interactions between atoms beloging to the same molecule . in the present formulation ,
however , the electrostatic energy in equation [ elec - final - qmmm ] arises from pairwise interactions of every ion with all the others , and those of intramolecular origin can not be easily individualized and excluded from the rest . a simple way to correct for
this overcounting could be to separately compute the coulomb interactions inside each molecule , and then substract it from the total energy . in this case
, such a correction would be : @xmath77 where @xmath78 is the number of water molecules in the mm region , and @xmath79 , @xmath80 and @xmath81 are the positions of the three atoms belonging to molecule @xmath82 . finally ,
combining all the contributions together , we compute the total energy as : @xmath83 = e_{es}[\rho ] + t_e[\rho ] + e_{xc}[\rho ] + e_{ps}^{nl } + e_{lj , im } + e_{lj , mm } + e_{bond } - e_{intra } \label{total - energy - hybrid}\ ] ] where @xmath84 considers the lennard - jones interactions within the mm region , and @xmath12 the intramolecular harmonic contributions between connected mm atoms ( equation [ ebond ] ) .
the atomic forces can be calculated for the qm and for the mm atoms as the derivative of the total energy , equation [ total - energy - hybrid ] , with respect to the ionic positions .
this leads to analytical forces in all cases . for the qm atoms
, there is no explicit dependence of @xmath85 $ ] and @xmath86 $ ] on @xmath87 , and therefore only three terms survive in the derivative : @xmath88 the former of these terms on the right hand side above can be developed as : @xmath89\ ] ] @xmath90 on the other hand , the lennard - jones contribution to the force is simply : @xmath91 \label{forces - qmmm-2}\ ] ] the contribution originating in the non local part of the pseudopotential energy is part of the standard qm implementation and will not be discussed in the present context .
the expression for the forces on the atoms belonging to the mm subsystem will have two terms in common with equation [ force - qm ] , plus the pure mm contributions , whose derivatives are straightforward : @xmath92
the potential energy curve of a water dimer , with one molecule in the qm domain and the other in the mm region , was calculated as a first test .
the geometry of the dimer was optimized for a series of oxygen - oxygen separations .
it is important to note that there are two inequivalent configurations in which this curve can be obtained , depending on whether the mm molecule plays the role of donor or acceptor of the hydrogen bond . in the water dimer
depicted in the inset of figure [ water - dimer ] , the hydrogen bond donor is the molecule on the left and the acceptor is on the right .
therefore , two curves can be obtained .
calculations were performed in a supercell of dimensions 36@xmath9318@xmath9318 bohr , to minimize the interactions of the water dimer with their periodic images .
the cell dimension is longer along the @xmath94 axis because this is the direction in which the potential energy is scanned .
the pbe approach to the dft exchange - correlation energy @xcite in combination with ultrasoft pseudopotentials @xcite were adopted to compute total energies and forces .
the kohn - sham orbitals and charge density were expanded in planewaves up to a kinetic energy cutoff of 50 and 200 ry , respectively .
an electronic mass of 400 a.u . was used to propagate the wavefunctions according to the car - parrinello scheme . before discussing the qm - mm results
, we will examine the curves provided by the pure qm ( dft ) and mm ( spc / fw ) methodologies .
these are plotted in figure [ water - dimer ] .
the difference in the description is readily apparent : the qm curve is shifted to a weaker interaction and to a longer distance at the minimum , with respect to the mm curve .
reported experimental values , of 2.98 for the oxygen - oxygen separation @xcite , and ranging from 3.6 to 5.2 kcal / mol for the magnitude of the interaction at the minimum energy geometry @xcite , are better reproduced by the quantum - mechanical model , which provides an optimized hydrogen - bond energy of 4.9 kcal / mol at 2.9 .
similar results have been obtained from previous dft calculations based on both localized and extended basis functions @xcite . on the other hand , the classical - mechanics water dimer interaction , of nearly 7.4 kcal / mol , is clearly above the available experimental data .
the reason for this overestimation is that the spc and spc / fw potentials , as most water force - fields , are parameterized to reflect the properties of the bulk liquid phase @xcite , where the molecular dipole moments are significantly enhanced with respect to the isolated molecule ( see next section ) . as a matter of fact
, the minimum of the mm curve , at nearly 2.7 , is coincident with the first peak of the oxygen - oxygen radial distribution function ( rdf ) for liquid water at room temperature .
the qm - mm potential energy curves , presented in figure [ water - dimer ] together with the results corresponding to the pure spc / fw and dft calculations , turn out to be quite interesting . as mentioned above ,
two configurations can be considered in this case , depending on the identity of the donor of the h - bond .
examination of the curves leads to the following observation : the qm - mm curve in which the acceptor is the mm molecule , roughly reproduces the spc / fw curve , whereas the qm - mm curve where this role is played by the qm molecule , is very close to the dft results .
in other words , the qm - mm curves are essentially reflecting the identity of the acceptor .
this is a meaningful result , understandable when we recall that most of the charge density involved in the bond , and therefore the polarization effect , corresponds to the oxygen atom .
the electron density associated with the h atom is much lower and localized , and there is not a major effect if this electron density is replaced by a bare pseudopotential .
ideally , both qm - mm curves should be identical . in practice , however , a discrepancy is immanent to all qm - mm models , since these curves are tied to the qm and mm hamiltonians , which necessarily provide different descriptions of the bond .
for the calculations in the condensed phase , we expect that the difference between the two kinds of interactions ( involving the qm water molecule as the donor or as the acceptor ) will be substantially attenuated .
the results in the next section , concerning the properties in the bulk , suggest that this is certainly the case .
car - parrinello molecular dynamics simulations were performed on a system of 64 water molecules , of which one was described quantum - mechanically and the rest classically .
the simulations were conducted at 300 k using the nose - hoover thermostat in a cubic box in periodic boundary conditions , with a density corresponding to 1 g/@xmath95 , and a time - length of 6 ps .
a planewaves basis of 25 ry and the pw91 exchange - correlation functional were employed @xcite .
the average dipole moment for a water molecule in the gas phase obtained from the car - parrinello dynamics is 1.81 d , very close to the experimental value of 1.86 d @xcite . in the system of 64 h@xmath0o molecules representing the bulk phase ,
the polarization exerted by the classical environment raises the dipole moment of the qm molecule to an average of 2.88 d. this number is in full agreement with reported estimates of 2.95 d from ab - initio simulations @xcite , or [email protected] d from the x - ray structure factor @xcite .
figure [ dipolemoments ] depicts the computed dipole moments of water as a function of time , both isolated and in the classical aqueous environment .
the dashed lines represent the experimental values .
such a good accord demonstrates that the polarization effect is finely accomplished by the qm - mm scheme .
figure [ fig - wat - freqs ] presents the vibrational frequencies of the water molecule in the gas and in the liquid phases , computed from the fourier transform of the time correlation function of the atomic velocities @xmath97 @xcite , @xmath98 depending on the system , on the vibrational mode , and on the electronic mass , ionic frequencies in car - parrinello dynamics may show redshifts of a few percent with respect to spectroscopic data @xcite . in the present case ,
the positions of the peaks fall between 100 to 200 @xmath99 below the experimental frequencies , consistently with previous car - parrinello simulations of h@xmath0o @xcite . for the isolated molecule , it is possible to recognize two groups of bands in figure [ fig - wat - freqs ] , corresponding to the stretching and bending modes , centered at 3500 and 1480 @xmath100 respectively . in the liquid state
these bands become broader and noisier , with an additional set of peaks below 1000 @xmath100 arising from librations . the first thing to note
is that the stretching frequencies shift to lower wavenumbers in the liquid , whereas the bending experiences the opposite trend .
this is the same behavior as observed from ir spectroscopy , where the stretching in the liquid is redshifted in about 300 @xmath100 , and the bending mode is blueshifted in nearly 50 @xmath100 @xcite .
our simulations in the liquid give peaks which are spread and too much splitted to establish unambiguously the magnitudes of these shifts ; however , considering the center of mass of the bands , it turns out that the shift of 300 @xmath100 in the stretching frequency is pretty much reproduced by the qm - mm model , while the change in the bending frequency appears overestimated by a factor of two .
these predictions for the spectral shifts in water are , from a quantitative point of view , of a quality comparable to that obtained from fully quantum - mechanical car - parrinello simulations @xcite . radial distribution functions for the model of 64 water molecules are displayed in figure [ fig - gdr ] .
classical molecular dynamics simulations with the spc - fw potential in pbc were performed in the same system with the lammps code @xcite .
the upper panel confirms that the electrostatic description of the mm atoms within the qm - mm approach , including the correction to the intramolecular coulomb forces , reproduces the dynamics dictated by the spc - fw force - field .
the subtle discrepancies between the classical and the qm - mm curves are attributable to differences in the length of the simulations , to the distinctive numerical implementations , and eventually to the presence of the qm water molecule .
the radial distribution function corresponding to the quantum - mechanical oxygen atom is presented in the lower panel of figure [ fig - gdr ] . in comparison to spc - fw ,
this curve is more structured , with maxima and minima appearing respectively above and below .
its shape does not seem to be converged , probably because insufficient sampling : note that in this case the rdf is built from a single water molecule out of 64 .
various studies of liquid water using the car - parrinello method with gga functionals and an electronic mass comparable to the one employed here , have found overstructured oxygen - oxygen rdfs at room temperature and pressure @xcite , suggesting that in these conditions this approach represents bulk water in a glassy or supercooled state .
data from one of these works is depicted in figure [ fig - gdr ] .
the radial distribution function of the qm oxygen atom in our qm - mm simulation turns out to be intermediate between those obtained from pure classical and pure quantum - mechanical car - parrinello molecular dynamics . at variance with other qm - mm schemes in the ppw framework , in which the classical region does not have to be included in the simulation box , in the present treatment
the mm atoms need to be contained in the supercell together with the qm atoms .
then , the amount of planewaves and the size of real space grids are the same or about the same as in a quantum - mechanical calculation with an equal number of total atoms , and therefore the qm - mm implementation does not involve any significant decrease in memory requirements . in spite of this , the reduction of the qm region cuts the quantity of kohn - sham states to be evolved in the car - parrinello dynamics , which may have a substantial impact on the overall computing time . as a matter of fact
, the computational effort in the quantum - mechanical car - parrinello scheme for a given unit cell size is approximately proportional to the number of atoms , with a slope larger than 1 , which tends to increase with system dimensions . as a consequence ,
the speedup achieved by the qm - mm approach is roughly linear with the replacement of qm by mm atoms .
this behavior is reflected in figure [ comp_time ] , which illustrates the relative decrease in computation time with the increase of the number of atoms represented classically , for the cases of 32 and 64 water molecules .
as expected , the gain in performance becomes more significant as the system grows bigger . in ab - initio modelling of solid interfaces in contact with a bulk liquid , the solvent fills a major fraction of the supercell , often representing between 1/2 to 2/3 of the total atoms . in the solid - liquid interface model for anatase ( 101 ) discussed in the next section , for instance ,
the number of classical atoms is 144 , out of a total of 264 . in this situation
, the qm - mm calculation turns out to be 5 times faster than a full qm simulation . in the case of a quantum - mechanical water molecule surrounded by 63 classical ones , examined in the previous section ,
the acceleration goes above one order of magnitude .
many of the most relevant applications of tio@xmath0 implicate the solid - liquid interface , which entails a serious challenge to first - principles modelling , given the need for extensive simulations to achieve an appropriate configurational sampling of the fluid phase .
thus , ab - initio molecular dynamics investigations considering the water liquid phase in contact with titania have been carried out only in a limited number of occasions , to address the rutile ( 110 ) @xcite and the anatase ( 101 ) and ( 001 ) surfaces @xcite . instead , there is a large number of dft studies which have examined stoichiometric and defective titania interfaces in the presence of just a few water monolayers , typically ranging from one to three @xcite . in this section
we illustrate the applicability of our hybrid quantum - mechanics molecular - mechanics scheme through the calculation of the water vibrational frequencies at the anatase ( 101 ) interface .
we aim at determining to what extent the explicit inclusion of the liquid phase affects the dynamical properties of the first adsorbed layers . to this end ,
results from qm - mm molecular dynamics simulations representing the liquid phase , are compared with those coming from qm simulations incorporating just one or two h@xmath0o monolayers .
the dft parameters concerning planewave basis , pseudopotentials , and exchange correlation functional , were the same as employed to describe the water bulk phase in the previous section .
the anatase ( 101 ) surface was represented using a 2@xmath932 supercell , containing six layers of tio@xmath0 units .
cell dimensions were [email protected]@xmath9322.67 @xmath101 for the qm calculations . in the qm - mm simulations
the @xmath102-parameter was extended to 30.23 to accomodate the aqueous phase , consisting of 16 qm plus 48 mm water molecules .
in particular , the first two water monolayers adjacent to the solid surfaces were modelled quantum - mechanically , to get the corresponding vibrational frequencies and to avoid a direct interaction between spc - fw water and titania .
molecular dynamics simulations were performed at 300 k with the nose - hoover thermostat , with sampling windows of 6 ps .
figure [ anatase - model ] displays the model structure employed in the qm - mm calculations of the solid - liquid interface .
anatase ( 101 ) surface in contact with a bulk water phase .
the water molecules were represented classically , excepting those forming the first and second adsorbed layers . ]
figure [ anat-1st ] shows the vibrational density of states corresponding to the water molecules directly adsorbed on the tio@xmath0 surface , computed through equation [ acf - vel ] .
the three panels compare three different coverages : monolayer ( top ) , bilayer ( center ) , and the liquid environment ( bottom ) . for the single layer ,
the stretching mode , appearing at around 3500 @xmath100 , is the strongest one , vaguely resembling the vibrational patterns of water in the gas phase . with the incorporation of a second layer ,
the intensities arising from librations and bending become larger than that associated with the stretching , which in turn moves slightly to the left .
this trend is similarly observed in the presence of the bulk liquid , where the librations dominate the spectra .
hence , in general terms it is observed that the vibrational behavior of the h@xmath0o molecules in the first layer , roughly shifts from gas - like to liquid - like as the degree of hydration is increased .
the main features in our computed spectra are in line with those obtained from neutron scattering experiments on water confined in anatase nanoparticles , where librations and bending are predominant , and the stretching absorption band is shifted to lower frequencies with respect to the gas phase @xcite .
the aqueous media has a major influence on the spectra of the second water layer .
figure [ anat-2nd ] shows that in the absence of the liquid environment , the stretching of the water molecules in the second layer is very weak in comparison with librations and bending .
the typical bulk - water features are recovered as the mm environment is included in the simulation .
the above results suggest that the internal degrees of freedom of a water molecule in a single layer on anatase ( 101 ) retain some of the character they have in the gas phase
. moreover , the sole inclusion of a second h@xmath0o layer appears to be enough to recreate , at least to some extent , the dynamics in the presence of a bulk aqueous phase .
this behavior can be tracked to the hydrogen bond network arising in each case .
for a single adsorbed layer , the h@xmath0o molecules are tightly bound to the five - coordinate ti atoms , with one or both of their hydrogen atoms forming relatively weak hydrogen - bonds with the bridging oxygen sites on the surface . on the other hand ,
an inspection of the trajectories in the presence of a second row of solvent molecules reveals that , most of the time , the molecules in the first layer are involved in hydrogen - bonds with at least another h@xmath0o molecule , either as a donor or as an acceptor . in the liquid phase the number of hydrogen bonds that can be formed , and the polarization effects , are even larger .
yet , the incorporation of a few water molecules beyond the first monolayer is sufficient to induce on the adsorbate , a dynamical behavior very close to the one corresponding to full hydration .
we have presented an approach to perform hybrid quantum - mechanics molecular - mechanics simulations with the car - parrinello method in the context of the pseudopotential - planewaves setting . at variance with other qm - mm implementations existing in planewaves codes , in the present approach the classical atoms are treated on the same footing as the quantum - mechanical ions , which naturally leads to periodic boundary conditions for the totality of the system .
thus , all qm and mm atoms need to be contained within the same real space grid determined by the simulation cell . as a consequence , the size of the mm region has an impact on the computational cost , and this method would not be convenient for extended systems where this region is extremely large , or exceeds by far the size of the qm part .
it turns out that , for a given unit cell , the scalability is approximately linear with the substitution of qm by mm atoms . in typical calculations of solid - liquid interface models involving a few hundred atoms , in which the solvent represents more than one half of the system , speedup factors above five can be attained with the present scheme .
we applied our implementation to the computation of the vibrational spectra of water adsorbed at the tio@xmath0 anatase ( 101 ) surface , at various coverages .
it was found that the presence of a second monolayer of water molecules is enough to mimick the effect of an aqueous environment on the vibrational frequencies of the first adsorbed layer .
this methodology seems particularly suited for molecular dynamics simulations in condensed matter systems including one or more fluid phases .
solutions , nanoconfined fluids , or solid - liquid and liquid - liquid interfaces , are all examples where this scheme could be extremely valuable .
we express our gratitude to ivan girotto and paolo gianozzi for precious help related to the structure and parallelization of the quantum - espresso code .
we also thank davide ceresoli for useful discussions .
this study has been partially supported by grants of the agencia nacional de promocion cientifica y tecnologica de argentina , pict 2012 - 2292 , and ubacyt 20020120100333ba .
we acknowledge csc - conicet for granting the use of the tupac hpc cluster , which allowed to perform most of the computations included in this work . | we introduce and discuss a hybrid quantum - mechanics molecular - mechanics ( qm - mm ) approach for car - parrinello dft simulations with pseudopotentials and planewaves basis , designed for the treatment of periodic systems . in this implementation
the mm atoms are considered as additional qm ions having fractional charges of either sign , which provides conceptual and computational simplicity by exploiting the machinery already existing in planewave codes to deal with electrostatics in periodic boundary conditions . with this strategy ,
both the qm and mm regions are contained in the same supercell , which determines the periodicity for the whole system .
thus , while this method is not meant to compete with non - periodic qm - mm schemes able to handle extremely large but finite mm regions , it is shown that for periodic systems of a few hundred atoms , our approach provides substantial savings in computational times by treating classically a fraction of the particles .
the performance and accuracy of the method is assessed through the study of energetic , structural , and dynamical aspects of the water dimer and of the aqueous bulk phase .
finally , the qm - mm scheme is applied to the computation of the vibrational spectra of water layers adsorbed at the tio@xmath0 anatase ( 101 ) solid - liquid interface .
this investigation suggests that the inclusion of a second monolayer of h@xmath0o molecules is sufficient to induce on the first adsorbed layer , a vibrational dynamics similar to that taking place in the presence of an aqueous environment .
the present qm - mm scheme appears as a very interesting tool to efficiently perform molecular dynamics simulations of complex condensed matter systems , from solutions to nanoconfined fluids to different kind of interfaces . |
as a potential candidate material for carbon - based electronics , graphene , a two - dimensional ( 2d ) honeycomb lattice of carbon atoms , has attracted much attention @xcite .
it is found that both the type ( electrons or holes ) and the number of carriers can be tuned by changing the gate voltage @xcite .
such tunability originates from the specific dispersion relation of the massless dirac fermions in graphene @xcite .
its transport properties are very special . there exists a universal maximal resistivity for all samples with the fermi level near the dirac point , independent of their shapes and mobility . the maximal resistivity @xcite implies the existence of channels of extended states at the thermodynamical limit , and the shape independence of the resistivity suggests that both the number and spatial configuration of these extended channels depend on the shape .
this behavior is certainly exotic as from the scaling theory all the states are localized in 2d systems @xcite .
the non - zero minimal conductivity has been investigated by several theoretical studies using different methods @xcite .
the minimum conductivity was correctly demonstrated , but as to the reason for it and as to its exact value being @xmath7 or @xmath8 there is no consensus at present .
et al . _ have found that in graphene the weak localization is strongly suppressed and they attributed this to mesoscopic corrugations of graphene sheets which can cause a dephasing effect similar to that of a random magnetic field @xcite .
theoretically , the influence of disorder on 2d electron gases on the honeycomb lattice was studied with a self - consistent born approximation @xcite .
_ have investigated the localized states near vacancies in graphene @xcite .
khveshchenko has carried out an analysis of quantum interference effects in disordered graphene and specified the conditions for the quantum correction to the conductivity to be positive , negative , and zero @xcite . by evaluating the dependence of the magnetoresistance of graphene on the relaxation rates from different sources
mccann _ et al .
_ showed that the warping disorder and the disorder due to atomically sharp scatterers have different effects on the magnetoresistance @xcite .
morpurgo and guinea investigated the effects of static disorder caused by curvature or topological defects and found that when the intervalley scattering time is long enough an effective time - reversal symmetry breaking can be induced @xcite .
aleiner and efetov showed two different types of logarithmic contributions to conductivity by describing the disorder with a renormalization - group equation @xcite . from a combination of mean - field and bosonization methods
altland pointed out that there are several mechanisms conspiring to protect the conductivity although disordered graphene is subject to the common mechanisms of anderson localization @xcite . from the linear response theory ziegler indicated that the transport properties in graphene are dominated by diffusion @xcite .
based on numerical study with finite - size kubo formula for doped graphene , nomura and macdonald found distinct differences between the cases of short - range and coulomb randomly distributed scatterers and speculated that the transport properties in graphene are related to the boltzmann theory for coulomb scatterers @xcite .
et al_. carried out transfer matrix studies of graphene in a magnetic field to investigate the disorder effect and phase diagram @xcite . louis _
et al_. calculated conductances for samples with edge disorder and obtained size independent conductivity as @xmath9 at @xmath0 @xcite . moreover , by adopting disorder model without intervalley scattering , ostrovsky _
et al_. constructed an effective field theory and showed that the system is at a quantum critical point with a universal value of the conductivity of the order of @xmath10 @xcite . with
a similar model titov demonstrated the impurity - assisted tunneling by using the transfer matrix approach @xcite .
et al_. numerically showed that the conductivity at the dirac point increases logarithmically with sample size in the absence of intervalley scattering @xcite . from numerical calculation nomura _
et al_. demonstrated that all states are delocalized even in the strong disorder regime if the intervalley scattering is excluded @xcite .
san - jose _
et al_. investigated the full transport statistics of graphene at low dopings without intervalley scattering and found identical scaling of all current cumulants @xcite . in this paper
we investigate the anderson localization of fermions at the dirac point in graphene with different types of disorder by using the standard finite - scaling analysis together with the transfer - matrix technique .
the dirac point corresponds to the band center ( @xmath0 ) of a tight - binding hamiltonian on the honeycomb lattice . in the real - space tight - binding representation
the disorder can be introduced with randomly distributed site energies , as called diagonal disorder , or with random hopping integrals named as off - diagonal ( od ) disorder .
it is known that the localization behavior at the band center of a bipartite lattice with od disorder is anomalous due to the chiral symmetry @xcite .
for instance , the decay of wavefunctions at the band center on 2d square lattice with od disorder is much weaker than exponential .
the honeycomb lattice is also bipartite , and there are other unusual features at the dirac point such as the massless dirac fermion dispersion relation .
so it is interesting to investigate the difference in the localization behavior at the dirac point between diagonal and off - diagonal disorders .
we derive the formula of transfer matrix for honeycomb lattice from which the numerical studies can be carried out to calculate interesting properties such as the rescaled localization length , the size and shape dependencies of the conductance .
the finite - size scaling analysis reveals that although all the states are localized in the system with diagonal disorder , in the case of od disorder , the states at the dirac point are delocalized . from the calculated shape dependence of the conductance at @xmath0 on an @xmath11 rectangular lattice we find that when all the hopping integrals are positive , the conductance is independent of @xmath5 , but if the distribution includes both the positive and negative hopping integrals , the conductivity becomes shape independent .
diagonal disorder corresponds to atomic scatterers such as doped impurities , while od disorder without sign changes of hopping integrals represents randomness of bond lengths caused by topological corrugations or warping in the sheets .
for graphene od disorder is more important than diagonal disorder owing to the ultrathin one - layer structure and the purity requirement in preparing the samples . in both types short - range disorder
can exist and the intervalley scattering is not excluded .
thus the obtained results should be different from those by using models without intervalley scattering due to different symmetry classes .
the sign randomness of hopping integrals can be produced by geometric or berry phase @xmath12 which is acquired by electron wavefunctions in every period of cyclic evolution around dirac point in parameter space .
the dirac point can serve as the degenerate point for geometric phase in slow " cyclic evolutions , such as in - plane vibrations of displacement between two sublattices , due to the specific dispersion relation and electron - lattice interaction .
the main features of the band structure in graphene can be well described by a tight - binding hamiltonian with one @xmath13 orbital per site on a honeycomb lattice @xcite : @xmath14 where @xmath15 ( @xmath16 ) creates ( annihilates ) an electron at site @xmath17 , @xmath18 is the energy level at site @xmath17 , @xmath19 denotes the nearest - neighbor ( nn ) sites , and @xmath4 is hopping integral between nn sites @xmath17 and @xmath20 . here , the spin indices are not explicitly included . in the absence of disorder ,
@xmath21 for all @xmath17 and @xmath22 for all @xmath23 , this hamiltonian leads to the 4-component dirac fermion dispersion relation near the dirac point @xmath0 .
the diagonal and off - diagonal disorder can be introduced by adopting random variables @xmath18 and @xmath4 satisfying distribution probabilities @xmath24 and @xmath25 , respectively .
these probability functions define strengths and types of disorder in the system .
diagonal disorder represents potential fluctuations due to impurities or due to randomness on substrate surface .
a most commonly used distribution probability for diagonal disorder is the square function @xmath26 where @xmath27 is the distribution width describing the strength of disorder . for od disorder , there are different distribution functions describing different types of randomness . if we consider slight fluctuations of bond lengths around their average value due to lattice distortions , the following distribution is suitable @xmath28 where @xmath29 ( @xmath30 ) is a measure of the bond - length fluctuations which make the nn hopping integrals randomly shift from @xmath31 . in @xmath32
all hopping integrals have the same sign as @xmath31 , reflecting that only changes of magnitude of @xmath4 are considered . in some cases
, however , @xmath4 not only fluctuate in magnitude , but also change in sign due to specific physical mechanisms .
one of these possible mechanisms is the geometric phase due to dynamical vibrations .
the on - plane displacement between two sublattices , denoted as @xmath33 , changes all the nn bond lengths and thus linearly couples with electron states in the regime of small @xmath34 . at a dirac point
( @xmath0 ) the states are doubly degenerate at @xmath35 .
the displacement @xmath36 linearly removes the degeneracy , leading to conical energy surfaces in the parameter space spanned by @xmath37 and @xmath38 with the degenerate point situated at the origin @xmath39 . owing to thermal excitations or even to zero - point oscillations ,
@xmath37 and @xmath38 vibrate and both of them are dependent on time with period @xmath40 where @xmath41 is the frequency of vibrations .
usually the amplitudes and relative phase of @xmath37 and @xmath38 vibrations are statistically random . for generic values of them ,
the track in the @xmath37-@xmath38 parameter space in a period is an ellipse encircling the degenerate point .
thus , according to the berry theorem , in each period of vibrations electron wavefunctions at dirac point will acquire a geometric phase @xmath12 in addition to the ordinary dynamical phase . at the same time in this period
the tunneling electron has traversed a path of length @xmath42 which equals @xmath43 nn bonds with @xmath44 and @xmath45 being the fermi velocity and bond length respectively .
this means that the electron states change sign in traversing every @xmath43 bonds .
such an effect occurs very naturally since the vibrations and electron transport are proceeding simultaneously . to include such an effect in a primary way
, we introduce the sign - changing probability of an nn hopping @xmath46 .
this probability can be estimated from the length by which the electron state changes sign due to the geometric phase , i.e. , @xmath47 .
with such a sign - change model the wavefunction can automatically change sign in traversing every length of @xmath42 which , averagely , contains only one hopping with opposite sign .
this is a special gauge in which the geometric phase acquired in a path of @xmath43 bonds is concentrated into one bond in this path .
this gauge is valid under the condition that the interferences between paths shorter than @xmath42 are not relevant .
as we will see below , the transport properties at dirac point in the case of od disorder ( including the sign changes ) are scale - invariant .
so the short - range processes are not important and the above condition can be satisfied .
this sign - change mechanism is different from the dephasing effect in the model of ref . where the distortions are fluctuating in the space ( the ripples ) but are static in time .
we introduce the following distribution of @xmath4 @xmath48 to account for the sign - change probability @xmath46 . when @xmath49 or @xmath50 , all the hopping integrals have the same sign and @xmath51 reduces to @xmath32 . at @xmath52
the probability of the sign changes is maximal .
below we set @xmath31 as energy units .
and @xmath53 are indices of columns ( shaded stripe along the @xmath54 direction ) and rows ( shaded stripe along the @xmath55 direction ) , and the overlapping area of two stripes is a basic cell where the lower and upper sites correspond to sublattices @xmath56 and @xmath57 , respectively . , width=634 ] the wavefunction of a fermion can be written as @xmath58 where @xmath59 is the amplitude on site @xmath17 . at a given energy @xmath60
the amplitudes are given by schrdinger equation @xmath61 .
the propagation properties of an electron on a finite lattice can be given by a relation between amplitudes on both sides .
this is usually formulated by the transfer - matrix method or by the green function .
we note that graphene lattice can be divided into two sublattices as distinguished by red and blue sites in fig .
1 . from the tight - binding nature of the hamiltonian and
the structure shown in fig .
1 it can be seen that for given @xmath60 all amplitudes can be iteratively calculated along the @xmath55 ( @xmath54 ) direction if the amplitudes in the light - cyan shaded column ( light - magenta shaded row ) are given . from this a transfer matrix along the @xmath55 or @xmath54 direction can be given .
we identify a site @xmath17 with three integers , @xmath62 , where @xmath63 and @xmath53 index the column and row , respectively , and @xmath64 specifies the sublattice which the site @xmath17 belongs to . in this notation
the iterative relation for amplitudes along the @xmath55 direction can be derived from schrdinger equation as @xmath65 @xmath66 for a system with finite width in the @xmath54 direction , one can define vectors for columns by @xmath67 , where @xmath68 is the number of cells in a column .
( [ 11 ] ) and ( [ 12 ] ) can be written in a matrix form : @xmath69 @xmath70 where elements of matrices are given by @xmath71 @xmath72 @xmath73 @xmath74 here , one may use the periodic or open boundary condition at the ends @xmath75 and @xmath76 .
we have the iterative relation for whole column @xmath77 where the transfer matrix is @xmath78 if there are @xmath79 columns in the @xmath55 direction , the total transfer matrix which gives the relation between amplitudes at two end columns is @xmath80 if we want to investigate the propagation properties in the @xmath54 direction , we have to establish the transfer matrix for rows . the relation between two adjacent rows is written as @xmath81 where @xmath82 s are vectors denoting rows with length @xmath68 , @xmath83 , and the elements of the transfer matrix are @xmath84 @xmath85 @xmath86 where @xmath87 @xmath88 the total transfer matrix for @xmath79 rows is @xmath89 for a finite @xmath2 system , there are @xmath68 propagating channels along the length @xmath79 .
the hermitian matrix , @xmath90 or @xmath91 , has @xmath92 eigenvalues .
these @xmath92 eigenvalues should be positive and come in inverse pairs due to the unitarity of the matrix @xcite .
the @xmath68 positive logarithms of eigenvalues , denoted as @xmath93 in an ascending order with index @xmath63 , are lyapunov exponents ( le ) reflecting the exponential decay of the corresponding channels . obviously , the logarithms of other @xmath68 eigenvalues are the negatives of @xmath94 .
the les will approach @xmath95 when the disorder vanishes .
the localization length is defined as the reciprocal of the smallest le @xmath96 .
the localization length plays a crucial role in the localization theory .
the zero - temperature conductance @xmath97 of a rectangular @xmath2 system in units of @xmath98 , with two spins being taken into account , can be evaluated using @xmath68-channel landauer formula @xcite @xmath99where the @xmath100 transmission matrix @xmath101 describes the transmission of electrons from one lead to the other in the longitudinal direction . here
we assume that the leads are connected to all the @xmath68 channels and they are perfect metals with band width much larger than that of graphene .
formula ( [ 6 ] ) can be expressed with the les as @xcite @xmath102 numerically , les can be calculated by using the standard method of gram - schmidt re - orthonormalization after each a few , say , ten steps of multiplication of transfer matrices @xcite .
this is equivalent to the diagonalization of @xmath103 or @xmath104 , but avoids terrible overflow and loss of precision on the computer . on the other hand , although the calculations of transfer matrices along the @xmath55 and @xmath54 directions are certainly different , there is no essential difference in the calculated results between these two directions .
this is owing to the fact that the dispersion relation is conic around the dirac point which is isotropic in the @xmath55 and @xmath54 directions .
below we will only present the results obtained for the transmission along the @xmath54 direction .
first we investigate the case of diagonal disorder , i.e. , @xmath105 , @xmath106 , and @xmath49 . the lowest le @xmath107 is calculated for very long ( @xmath108 ) strip of width @xmath68 . the rescaled localization length is defined as @xmath109 . for given energy and disorder strength ,
the localization behavior of the fermions can be determined by the scaling dependence of @xmath110 : if @xmath110 increases with @xmath68 , the states are extended .
for localized states @xmath110 is decreased by increasing @xmath68 .
the @xmath68-independence of @xmath110 corresponds to the fixed point separating regions of extended and localized states . in fig .
2(a ) , we plot the @xmath68 dependence of @xmath110 for states at @xmath0 in systems with different strengths of diagonal disorder . for all investigated strengths of diagonal disorder
, the rescaled localization length always decreases with increasing width .
the large fluctuations in the curve of @xmath111 are due to numerical errors in the gram - schmidt re - orthonormalization procedure which become serious in the case with @xmath112 .
this means that the states at the dirac point are localized by tiny diagonal disorder . in fig .
2(b ) , we plot the @xmath68-dependence of @xmath110 for a given strength of diagonal disorder but different energies .
although all states are localized , the scaling behavior is much different between the states at dirac point and the other states : the decreasing slope of the former is much larger than that of the latter .
this implies that the states at dirac point are more easily localized than other states by introducing the diagonal disorder in the thermodynamical limit , in spite of their larger localization length in smaller systems
. this may be due to the small density of states at @xmath0 and the breaking of the chiral symmetry by diagonal disorder .
the localization behavior of the states is consistent with the common theories of the anderson localization in 2d . versus @xmath68 at the dirac point ( @xmath0 ) for different strengths of disorder .
( b ) @xmath110 versus @xmath68 for a given strength of disorder and various energies .
, title="fig:",width=332 ] versus @xmath68 at the dirac point ( @xmath0 ) for different strengths of disorder .
( b ) @xmath110 versus @xmath68 for a given strength of disorder and various energies .
, title="fig:",width=332 ]
differently from diagonal disorder , off - diagonal disorder does not break the chiral symmetry at dirac point , leading to an essential difference in the localization .
the scaling analysis of the rescaled localization length and the conductance at dirac point is shown in fig .
3 by using the distribution function @xmath113 in eq .
( [ p01 ] ) .
this corresponds to the case of randomly warped or corrugated lattice where bond lengths are randomly shifted from the perfect value . from fig .
3(a ) we can see that the rescaled localization length is independent of @xmath68 for all disorder strengths .
this implies that the localization length at the thermodynamical limit is infinite , corresponding to delocalized states at @xmath0 .
the conductance of an @xmath100 system calculated from eq .
( [ cond ] ) is also independent of @xmath68 for all investigated strengths of off - diagonal disorder [ see fig .
so the conductance at the thermodynamical limit can be determined from these size independent values .
the dependence of @xmath97 on the disorder strength is shown in the inset of fig .
for relatively weak off - diagonal disorder ( @xmath114 ) the conductance at dirac point is about @xmath6 , corresponding to the quantum conductance of 4 channels ( 2 spin channels and 2 equivalent channels from the chiral symmetry at @xmath0 ) .
in fact , in a system with off - diagonal disorder , the calculated lyapunov exponents at @xmath0 can be grouped into pairs , i.e. , @xmath115 for integer @xmath63 in the range between 1 and @xmath116 , corresponding to two equivalent channels with the chiral transformation .
these channels have the localization length proportional to the width @xmath68 . at the thermodynamical limit
they become infinite . by increasing the disorder strength from @xmath117 towards its maximum value @xmath118
, these channels are still delocalized at the thermodynamical limit but the conductance rapidly drops , as can be seen from the inset of fig .
3(b ) . system ( b ) at the dirac point ( @xmath0 ) in the presence of off - diagonal disorder .
the inset of ( b ) shows the dependence of @xmath97 on the strength of off - diagonal disorder @xmath29 . , title="fig:",width=332 ] system ( b ) at the dirac point ( @xmath0 ) in the presence of off - diagonal disorder .
the inset of ( b ) shows the dependence of @xmath97 on the strength of off - diagonal disorder @xmath29 .
, title="fig:",width=332 ] the delocalization behavior at @xmath0 in graphene with the off - diagonal disorder is not consistent with the usual localization behavior of 2d disordered systems .
one may attribute this inconsistency to the chiral symmetry at @xmath0 which is conserved in off - diagonal disorder but destroyed in diagonal disorder .
so the localization at energies @xmath119 , where the chiral symmetry no longer exists , could be expected . in fig .
4 , we plot the scaling behavior of the rescaled localization length and the conductance of an @xmath100 system for different energies . except for the curves of @xmath0 , both the rescaled localization length and the conductance decrease by increasing @xmath68 in the general trend if fluctuations of some curves are ignored , suggesting the usual localization behavior at @xmath119 . from fig .
4(a ) , the rescaled localization length is generally decreased by increasing @xmath60 from @xmath0 for a fixed @xmath68 .
this means that in off - diagonal disorder the states of energies farther apart from dirac point are more localized . from fig .
4(b ) , however , for a given value of @xmath68 the conductance at larger @xmath120 has larger values .
this can be explained by the fact that there are more electron channels at larger @xmath120 in graphene which can contribute to the conductance if @xmath68 is not very large
. system ( b ) for different energies in the presence of off - diagonal disorder . , title="fig:",width=332 ] system ( b ) for different energies in the presence of off - diagonal disorder .
, title="fig:",width=332 ] by using a scaling transformation @xmath121 with appropriate value of scaling factor @xmath122 for every energy @xmath60 , the curves in fig .
4(a ) for different energies can merge together . in fig .
5 we plot the common dependence of @xmath110 on @xmath123 for various energies . the scaling factor @xmath122 as a function of @xmath60 is shown in the inset which can be fitted by a simple function @xmath124 with prefactor @xmath125 . since @xmath122 is just the localization length at the thermodynamical limit , this scaling behavior further ensures the delocalization at @xmath0 and describes how the localization length diverges when @xmath126 . from the dirac dispersion relation
the wave length of electrons is also inversely proportional to @xmath120 .
this suggests that in the case of off - diagonal disorder the localization length and the electron wave length have the same @xmath60 dependence and can be regarded as the same length scale .
we also notice that although the states at @xmath0 are delocalized , there is no real metallic region with finite range of @xmath60 .
this is certainly different from the results in refs . where the disorder models free of intervalley scattering are adopted . on the scaled width @xmath127 for various energies .
inset : @xmath122 as a function of @xmath60 ( circle symbols ) which is fitted by function @xmath128 ( red curve ) .
, width=332 ] the different localization behaviors between @xmath0 and @xmath119 can be more clearly depicted by the spatial distributions of wavefunctions as shown in fig .
6 . as can be seen from fig .
6(a ) , the delocalized state at @xmath0 is more likely to be a cantor - set like critical state , rather than a real metallic state .
such states may provide propagation paths for current from one edge to the other . on the other hand , the state at @xmath129 shown in fig .
6(b ) has significant amplitudes only in some isolated islands , corresponding to typical localization . and ( b ) @xmath129 in a @xmath130 system in off - diagonal disorder of strength @xmath131 . , title="fig:",width=332 ] and ( b ) @xmath129 in a @xmath130 system in off - diagonal disorder of strength @xmath131 .
, title="fig:",width=332 ] in the case of off - diagonal disorder there are 4 delocalized channels at dirac point , giving conductance @xmath6 .
this leads to almost shape - independent conductance and shape - dependent conductivity .
the situation is similar to the case of perfect lattice investigated by tworzydlo _
et al_. in ref . .
following the procedure proposed in ref .
, we calculate the conductivity from the conductance of a @xmath2 rectangular sample .
the obtained shape dependence of the conductivity is shown in fig .
this behavior is consistent with the nature of the quantum transport in 2d , but is inconsistent with the experiments where the conductivity is shape independent .
this means that to explain this experimental result we have to invoke some mechanism which could convert the quantum transport to the classical - like transport .
rectangular system with a given length at @xmath0 for different strengths of off - diagonal disorder .
, width=332 ]
the quantum transport processes could be drastically changed if the phases of wavefunctions become uncertain .
this may happen due to the geometric phase which can be acquired by electrons during a cyclic evolution of slow " degrees of freedom around a conic point .
the additional geometric phase of @xmath12 corresponds to sign changes of the wavefunction during the motion of tunneling electrons . in the diffusion regime , if the sign changes randomly happen during the spatial motion , they can be modeled by a random distribution of positive and negative signs of hopping integrals . in probability @xmath51 of eq .
( [ signn ] ) we include both the sign randomness and the magnitude randomness of the hopping integrals which are specified by parameters @xmath46 and @xmath29 , respectively .
first we set @xmath106 and focus on the effect of sign randomness . in fig .
8 we plot the scaling behavior of the rescaled localization length and the conductance of an @xmath100 system for different values of @xmath46 .
similarly to the case of off - diagonal disorder , the states at the dirac point in the presence of the sign randomness are also delocalized in the sense of finite - size scaling analysis . differently from the behavior in the off - diagonal disorder shown in fig .
3 , effects of sign randomness on the rescaled localization length and on the conductance are opposite : @xmath132 in @xmath52 is smaller than that in @xmath133 , but @xmath97 in @xmath52 is larger than that in @xmath133 .
this is strange because @xmath110 corresponds to the channel with the largest localization length which usually provides the leading contribution to the conductance .
the only explanation for this is that the other channels also give significant contributions in the case of sign randomness .
an expectable consequence of this is a new shape dependence of the conductivity , since the number of the tunneling channels is no longer restricted .
system in the presence of sign randomness of hopping integrals obeying probability @xmath51 with different values of @xmath46 .
, width=332 ] the shape dependence of the conductivity can be investigated by calculating the conductance of @xmath2 rectangular samples .
the obtained conductivity as a function of ratio @xmath3 for various sizes is shown in fig .
9 . except for small values of @xmath3 ,
the conductivity has values near @xmath6 , independent of @xmath3 and @xmath68 , in consistence with the experimental findings .
surprisingly , this value is even larger than @xmath134 obtained in ballistic graphene @xcite in spite of the disorder introduced by the sign randomness of hopping integrals . to get insight into this anomalous feature ,
we adopt the average spacing between successive pairs of lyapunov exponents denoted as @xmath135 , i.e. , @xmath136 .
notice that in this case the lyapunov exponents are grouped into pairs , corresponding to two equivalent channels from the chiral symmetry . by keeping only the leading exponential term of @xmath137 , eq .
( [ cond ] ) can be approximately rewritten as @xmath138 in units of @xmath139 . in a perfect system of width @xmath68 and using the periodic boundary condition , the spacing @xmath140 for states at @xmath0 is @xmath141 due to the dirac fermion dispersion relation @xcite .
one may reasonably suppose that @xmath142 is still held in the case of od disorder .
then , for @xmath143 , the conductivity of an @xmath2 rectangular sample can be calculated as @xmath144 so for the perfect lattice @xmath145 in units of @xmath139 @xcite . in the case of sign randomness ,
the values of @xmath146 are in the range @xmath147 $ ] as shown in the inset of fig .
9 . together with the factor @xmath148
, the obtained conductivity may be @xmath13 times of that of the perfect lattice .
thus , the @xmath13 factor difference in the conductivity is not trivial and reflects an essential change of the fermion properties due to the phase uncertainty . for rectangular sheets with different widths @xmath68 .
inset in the lower panel : distribution function @xmath149 of @xmath150 .
the green dashed line indicates the @xmath151-function distribution at @xmath152 in the case of perfect sheet . , width=332 ] it is interesting to investigate how the localization behavior changes in the presence of the magnitude randomness of hopping integrals .
the scaling behaviors of the rescaled localization length and the conductance of an @xmath100 system with both nonzero @xmath46 and @xmath29 are shown in fig . 10 .
it can be seen that the states at @xmath0 are still delocalized and the opposite effects on the rescaled localization length and on the conductance are more evident than those shown in fig .
8 . in fig .
11 we display the shape dependence of the conductivity of the @xmath2 rectangular samples for @xmath52 and different nonzero values of @xmath29 .
similarly to the case of @xmath153 , the conductivity is still shape independent except for small values of @xmath3 .
this means that the sign - randomness induced shape independence of the conductivity is robust against the warping or corrugation disorder which causes randomness of magnitudes of hopping integrals .
it can also be seen that the value of the conductivity is slightly reduced from @xmath6 by increasing the warping or corrugation disorder . for reasonable strength of the warping disorder ( @xmath154 ) , however , the conductivity is above @xmath155 , in the range of measured values in experiments .
system in the presence of both sign randomness and magnitude randomness of hopping integrals obeying probability @xmath51 .
, width=332 ] for rectangular sheets with different widths @xmath68 in the presence of both sign randomness and magnitude randomness of hopping integrals . , width=332 ] it is also interesting to investigate the scaling behavior of the statistics of the minimum conductivity . in fig .
12 we plot the statistical distribution of the conductivity at @xmath0 for various system sizes and for given values of @xmath3 , @xmath29 , and @xmath46 .
we note that the distribution of the conductivity is gaussian - like , and both the average and the distribution width are almost scaling invariant .
this is different from the statistics of a single lyapunov exponent in usual disordered system for which the distribution width is @xmath156 dependent .
this difference originates from the fact that in the present system the conductivity is contributed not only from the channel with the smallest le but also from other channels . as a consequence the statistics of the conductivity given by eq .
( [ ooo ] ) is mainly determined by the distribution of @xmath157 rather than by the distribution of a single le .
it can be easily verified that in the present case the series of successional les times @xmath68 , @xmath158 are scaling invariant . thus the statistics of @xmath157 is also scaling invariant , leading to the scaling invariant statistics of the minimum conductivity . for various system sizes in od disorder with sign randomness .
the values of @xmath3 , @xmath29 , and @xmath46 are given in the figure .
the statistics is taken from 500 random configurations .
, width=332 ]
we investigate the effect of different types of disorder on the transport properties of electrons at dirac point in graphene . from the finite - size scaling analysis together with the transfer - matrix technique
, it is found that all states are localized in the presence of the diagonal disorder , in consistence with the main conclusions of the scaling theory for 2d . in the case of off - diagonal disorder , however , the states at @xmath0 are delocalized .
this delocalization , inconsistent with the scaling theory and only existing at dirac point ( @xmath0 ) , is mainly due to the specific chiral symmetry which is conserved in off - diagonal disorder but destroyed by diagonal disorder .
the obtained conductance at the thermodynamical limit shows two transparent chirally symmetric quantum channels at @xmath0 .
these channels , however , could not produce shape independent conductivity observed in experiments .
it can be obtained only by introducing the sign randomness of hopping integrals .
we investigate the scaling behavior of the rescaled localization length and the conductance in the presence of both sign randomness and magnitude randomness to show the existence of delocalized states at @xmath0 .
especially , the calculated conductivity of a rectangular sample is shape independent .
as the sign randomness of the hopping integrals corresponds to some additional uncertain phases , the obtained results reflect a general trend of transition from quantum to classical transport in dephasing processes .
although the calculated conductivity is in consistence with the experimental findings , the mechanism for the phase uncertainty still needs to be identified by future investigations .
this work was supported by the state key programs for basic research of china ( 2005cb623605 and 2006cb921803 ) , and by national foundation of natural science in china grant nos .
10474033 , 10704040 , and 60676056 . | anderson localization of electron states on graphene lattice with diagonal and off - diagonal ( od ) disorder in the absence of magnetic field is investigated by using the standard finite - size scaling analysis . in the presence of diagonal disorder
all states are localized as predicted by the scaling theory for two - dimensional systems . in the case of od disorder , the states at the dirac point ( @xmath0 )
are shown to be delocalized due to the specific chiral symmetry , although other states ( @xmath1 ) are still localized . in od disorder
the conductance at @xmath0 in an @xmath2 rectangular system at the thermodynamical limit is calculated with the transfer - matrix technique for various values of ratio @xmath3 and different types of distribution functions of the od elements @xmath4 .
it is found that if all the @xmath4 s are positive the conductance is independent of @xmath5 as restricted by 2 delocalized channels at @xmath0 .
if the distribution function includes the sign randomness of elements @xmath4 , the conductivity , rather than the conductance , becomes @xmath5 independent .
the calculated value of the conductivity is around @xmath6 , in consistence with the experiments . |
in condensed matter physics , an emergent relativity at low energies appears for systems with two stable fermi points , see , _ e.g. _ , the textbook ref .
[ ] . this is the case of undoped graphene , a one - atom thick layer of graphite , see , _
e.g. _ , ref .
[ ] for a review , where the quasiparticle spectrum is dirac - like and massless at low - energies @xcite . in ref .
[ ] a renormalization group approach indeed revealed the existence of an infrared lorentz invariant fixed point for graphene . approaching this fixed point ,
the fermi velocity , @xmath0 , flows to the velocity of light , @xmath1 , while the fine structure constant of graphene , @xmath2 , which is of order one , flows to the usual fine structure constant , @xmath3 .
moreover , while electrons in graphene are confined to a three - dimensional space - time , @xmath4 , interactions between them are mediated by four - dimensional photons , @xmath5 .
the lorentz invariant fixed point may therefore be effectively described by a massless relativistic quantum field theory ( qft ) model whereby @xmath6-dimensional electrons interact via a long - range fully retarded potential .
such a model belongs to the class of reduced quantum electrodynamics ( rqed ) [ ] , or rqed@xmath7 , and corresponds to rqed@xmath8 in the case of graphene .
following ref .
[ ] , the latter is also sometimes referred to as pseudo - qed , see , _
e.g. _ , ref .
[ ] . in the general case , rqed@xmath7 is a relativistic qft describing the interaction of an abelian @xmath9 gauge field living in @xmath10 space - time dimensions with a fermion field living in a space - time of @xmath11 dimensions .
the case where @xmath12 corresponds to usual qeds , see the textbooks refs .
the reduced case corresponds to @xmath13 where the fermion field is localized on a @xmath14-brane .
motivated by potential applications to condensed matter physics as well as interest in branes we focus on the computation of radiative corrections in a general model of rqed@xmath7 .
the computations are based on sophisticated methods devoted to the exact evaluation of multi - loop feynman diagrams in relativistic qfts ; these methods include , _
e.g. _ , the gegenbauer polynomial technique @xcite , integration by parts @xcite ( ibp ) , and the method of uniqueness .
@xcite in refs .
[ ] multi - loop corrections in a general theory of rqed@xmath7 were computed with a special emphasis on electromagnetic current correlations and the link between rqed@xmath8 and the ultrarelativistic limit of graphene was made .
in particular , the two - loop interaction correction coefficient to the polarization operator of graphene in the ultrarelativistic limit was derived and found to be small in qualitative agreement with theoretical @xcite as well as experimental @xcite results in the non - relativistic regime .
these results were used in ref . [ ] to show that the optical conductivity of graphene in the ultrarelativistic limit has a relative deviation which is within experimental uncertainty @xcite and of the order of one percent with respect to the non - interacting value . this striking qualitative agreement with the non relativistic limit together with the efficient tools available to tackle the relativistic limit suggest the crucial importance of fully exploring the properties of the lorentz invariant fixed point .
we pursue this task in the present paper by computing multi - loop corrections to the fermion propagator and the corresponding anomalous scaling dimension of the fermion field in rqed@xmath15 .
a peculiarity of graphene is that the fermion field gets renormalized but not the electric charge , see ref .
a similar feature was found for rqed@xmath8 in ref .
[ ] where the anomalous dimension of the fermion field was computed at one - loop . in the following we extend these results to two loops . from the field theory point of view , the main technical difficulty involves the computation of some peculiar two - loop massless propagator diagrams with two non - integer indices on non - adjacent lines ( see eq .
( [ complicatedg ] ) below ) .
presently , no explicit analytical expression for such diagrams is known .
it turns out that , as will be shown below , repeated use of some well chosen ibp identities allows us to express each complicated , and eventually divergent , diagram as a sum of primitive ( or recursively one - loop ) diagrams plus a complicated but convergent diagram which is multiplied by a factor @xmath16 . in the limit , @xmath17 , the complicated part does not contribute which solves the problem in the case of rqed@xmath15 .
the paper is organized as follows . in sec .
[ sec : rqed ] , we review the basics of massless rqed at one - loop . in sec . [
sec : sigma ] , we compute the two - loop fermion self - energy in rqed@xmath7 using ibp identities and show that , for @xmath17 , it can be expressed only as a function of recursively one - loop diagrams . in sec .
[ sec : fermidim ] , we compute the anomalous scaling dimension of the fermion field and the renormalized fermion propagator at two - loop , in an arbitrary gauge in the general case of rqed@xmath15 . the special case of rqed@xmath8
is then explicitly considered as well as , for completeness , the cases of qed@xmath18 and qed@xmath19 . in sec .
[ sec : conclusion ] we summarize our results and conclude .
finally , app .
[ app - gfuncs ] contains the expansion of some master integrals entering the two - loop self - energy and app .
[ app : g(alpha , beta ) ] presents general formulas , valid beyond ibp relations , for diagrams with two non - integer indices . in the following , we work in units where @xmath20
we consider a general model of massless rqed@xmath7 described by [ ] : s _ = & & ^d _ x _ , [ rqed ] + _ = & & ( x ) i ^_ed__e ( x)^(d_-d_e)(x ) - + & & -f___f^ _ _ - ( _ _ a^_)^2 , where @xmath21 is the covariant derivative , @xmath22 is the field stress tensor of the gauge potential @xmath23 and @xmath24 is a gauge fixing parameter . in eq .
( [ rqed ] ) , on the one - hand , the index @xmath25 runs over the @xmath11-dimensional space - time in which fermions are localized : @xmath26 , where zero index corresponds to time . on the other hand ,
the index @xmath27 runs over the @xmath10-dimensional space - time of the gauge field : @xmath28 , where we assume that @xmath29 .
the minimal coupling of the gauge field to the fermion current , @xmath30 , is restricted to the reduced matter space from which we deduce the expression of the fermion current : j^(x ) = \ { ll ( x ) ^(x ) ^(d_-d_e)(x ) & = 0,,d_e-1 , + 0 & = d_e,,d_-1 . + . [ fc ]
simple dimensional analysis then shows that two epsilon parameters are naturally associated with the two dimensions @xmath10 and @xmath11 .
indeed , for a general rqed@xmath7 , the dimensions of the fields and electric charge may be written as : = 1 - _ , = - _ e - _ , = _ , [ dims ] where the two epsilon parameters read : _
= , _ e = , [ params ] and will be shown below to play a crucial role in setting up a dimensional regularization scheme for rqed .
conversely , eq . ( [ params ] ) yields the relations : d_= 4 - 2 _ , d_e = 4 - 2_e - 2_. [ d - params ] as can be seen from eq .
( [ dims ] ) the dimension of the coupling constant is entirely determined by the space - time dimension of the gauge field , @xmath10 , or , equivalently , @xmath31 ; for a four - dimensional gauge field ( @xmath32 ) the coupling is dimensionless so that , at least at a classical level , the theory is scale invariant .
this suggests that rqed@xmath15 are renormalizable qfts .
following ref .
[ ] , let s recall that the superficial degree of divergence ( sdd ) of a general rqed@xmath7 diagram reads : , eq .
( [ sdd ] ) yields : @xmath33 .
the latter exactly corresponds to the result of ref .
[ ] , see discussion and equations below eq .
( 2.20 ) in this paper , for qed@xmath19 in the large-@xmath34 limit . ]
d = d_e + v - n_-n_e , [ sdd ] where @xmath35 is the number of vertices , @xmath36 the number of external gauge lines and @xmath37 the number of external fermion lines . from eq .
( [ sdd ] ) we see that for @xmath32 the sdd does not depend on the number of vertices whatever value @xmath11 takes ; this confirms the renormalizability of the theory .
a peculiar fact of reduced theories is that , while the fermion self - energy and fermion - photon vertex have the same degree of divergence as in usual qed@xmath18 ( they effectively diverge logarithmically ) , the photon self - energies of rqed@xmath8 and rqed@xmath38 are finite .
this is summarized in tab .
[ tab : sdd ] which displays the degrees of divergence ( superficial and effective ) of the three most divergent amplitudes in qed@xmath18 , rqed@xmath8 and rqed@xmath38 . as a consequence , while the fermion field acquires an anomalous dimension , the coupling constant does not renormalize .
.superficial degree of divergence ( and effective degree of divergence ) of the three most divergent amplitudes in rqed@xmath15 for @xmath39 ( qed@xmath18 ) , @xmath40 and @xmath41 .
[ cols="<,^,^,^",options="header " , ] the above dimensional analysis can be made quantitative by setting up a perturbative approach to rqed . as can be seen from eq .
( [ rqed ] ) the free massless fermion propagator is the usual one and reads : s_0(p_e ) = , [ fermion - prop0 ] where @xmath42 lies in the reduced matter space and @xmath43 is a convergence factor that will often be omitted in the following in order to simplify notations . on the other hand , because we are interested in the properties of the reduced system , we may integrate over the @xmath44 bulk gauge degrees of freedom .
this yields an effective free gauge field propagator reading : _ 0^(q ) = ( g^ - ) , [ gauge - field - prop0 ] where , now , @xmath45 lies in the reduced matter space and @xmath46 was defined in eq .
( [ params ] ) .
moreover , we see that the gauge fixing parameter is affected by the integration of the bulk gauge modes : @xmath47 where @xmath48 .
the two commonly used relativistic gauges are the feynman and landau gauges which are defined as : [ gauges ] = 0 & & a=1 , [ feynmangauge ] + = 1 & & a=0 .
[ landaugauge ] in the following we shall mainly work in an arbitrary gauge . in the case of rqed@xmath8
for which @xmath49 , the effective photon propagator of eq . ( [ gauge - field - prop0 ] ) has a square root branch cut _ 0^(q ) = ( g^ - ) .
[ gauge - field - prop0-rqed43 ] this momentum dependence is responsible for the appearance of feynman diagrams with non - integer indices and is a major source of technical difficulty that we shall discuss in details in the next sections . as already noticed in ref .
[ ] , a similar momentum dependence can be found for qed@xmath19 in the large-@xmath34 limit where @xmath34 is the number of fermion species .
the reason is the `` infrared softening '' of the photon propagator of qed@xmath50 for large @xmath34 @xcite . indeed , using a non - local gauge fixing term , @xmath51 , as done in ref .
[ ] and justified in the general case in ref .
[ ] , the dressed photon propagator together with the one - loop polarization operator of massless qed@xmath19 , read : d^(q ) = -i,_1(q^2 ) = - , [ gauge - field - prop - qed3 ] where , for @xmath52 , the usual landau gauge propagator is recovered , the lower index @xmath53 on @xmath54 refers to the order of perturbation theory and @xmath55 has dimension of mass in super - renormalizable qed@xmath19 . in the @xmath56 approximation ( @xmath57 and @xmath58 with @xmath59 fixed ) and focusing on the infrared limit , @xmath60 , the dressed photon propagator of eq .
( [ gauge - field - prop - qed3 ] ) becomes : d^(q ) = ( g^ - ) , [ gauge - field - prop - qed3 - 2 ] and , similarly to eq .
( [ gauge - field - prop0-rqed43 ] ) , behaves like @xmath61 instead of the usual @xmath62 . finally , the effective free photon propagator of rqed@xmath38 , which has an exponent @xmath63 , has even a softer momentum dependence .
the latter is similar to the momentum dependence of the dressed photon propagator of qed@xmath64 at large @xmath34 . in the case of rqed@xmath38
, however , the logarithmic divergence of the effective photon propagator translates into the appearance of a pole in the gamma function of eq .
( [ gauge - field - prop0 ] ) .
this case requires some further regularization such as , _ e.g. _ , giving a finite width to the 1-brane @xcite . in the following ,
we shall mainly focus on the 2-brane case described by rqed@xmath8 .
the similarity between the reduced versions of qed and its large @xmath34 limit will be a subject of our future investigations .
feynman rules for massless rqed . ] with the feynman rules of massless rqed@xmath7 summarized on fig .
[ fig : feynmanrules ] , perturbation theory is then implemented in the usual way by solving dyson equations for the fermion and photon propagators and computing the various self - energies .
focusing for simplicity on the fermion propagator , which is central to the present work , the corresponding dyson equation reads : s(p ) = s_0(p ) + s_0(p ) ( -i ( p ) ) s(p ) , where @xmath65 is the fermion self - energy .
the solution to this equation in the case of massless fermions can be written as : -i s(p ) = , ( p ) = _ v(p^2 ) .
[ fermion - prop+sigma ] the effective divergence appearing in parentheses in tab .
[ tab : sdd ] actually corresponds to the divergence of @xmath66 .
such a self - energy is divergent in all rqeds as indicated in the table .
the divergence is related to loop integrals depending on @xmath11 . following refs .
[ ] we use dimensional regularization and work with an arbitrary @xmath11 .
the latter may be expressed as a function of the two epsilon parameters which naturally appeared at the level of eq .
( [ params ] ) .
the parameter @xmath67 takes into account the brane - like nature of the system . for fixed @xmath67 self - energies and related propagators
take the form of laurent series in @xmath31 . working in an improved minimal subtraction scheme ( @xmath68 )
we may then absorb the singular part of the series ( in @xmath69 ) in renormalization constants which relate the bare fields and parameters of the theory to renormalized ones : = z_^1/2_r , a = z_a^1/2 a_r , a = z_a a_r , e = z_^1/2e_r , [ ren_constants ] where @xmath70 , @xmath71 and @xmath72 are the dimensionless renormalization constants .
moreover , in the @xmath68 scheme we also define a dimensionless renormalized coupling constant @xmath73 via the equation : = ^-2_z_^-1(())e^-_e _ , [ ren_coupling ] where @xmath74 is euler s constant and the @xmath75 factor compensates for the dimension of @xmath55 .
equivalently , from eq .
( [ ren_coupling ] ) , the bare coupling constant @xmath76 can be expressed via the renormalized coupling constant @xmath77 . from the above arguments ,
the relation between the bare and renormalized fermion propagators reads : s(p ) = z_((),a_r ( ) ) s_r(p ; ) , [ mpropr ] where all singularities are in @xmath70 and @xmath78 , the renormalized fermion propagator , is finite .
similarly , the relation between the bare and renormalized effective photon propagators reads : ^(q ) = z_a ( ( ) ) _ r^(q ; ) , where all singularities are in @xmath71 and @xmath79 , the renormalized effective photon propagator , is finite .
finally , one may introduce another constant , @xmath80 , for the vertex renormalization : @xmath81 where @xmath82 is finite . because both the renormalized coupling and vertex are finite there is a constraint among the constants : z_= ( z_z_)^-2 z_a^-1 .
[ zs - constraint ] furthermore , as in usual qeds , the ward identity : z_z_= 1 , [ ward ] is satisfied in a general rqed implying that coupling renormalization is entirely due to gauge field renormalization : @xmath83 . ) , in rqed follows the same steps as for usual qeds , see , _ e.g. _ , the textbooks refs .
the reason is that the textbook proof does not depend on the index of the gauge - field propagator which may then be arbitrary . ]
the beta function can then be expressed as a function of the anomalous scaling dimension of the gauge field as follows : & & ( ( ) ) = = -2_+ _ a ( ( ) ) , + & & _ a ( ( ) ) = .
[ beta+gammaa ] these functions together with the renormalized photon propagator of rqed were considered in refs .
[ ] where it was shown , in accordance with dimensional analysis , that : @xmath84 in rqed@xmath15 ( @xmath85 . in the following we will compute the renormalization constant @xmath70 and deduce the anomalous scaling dimension of the fermion field : _ ( ( ) , a_r ( ) ) = .
[ asd ] differentiating eq .
( [ mpropr ] ) with respect to @xmath86 and taking into account eq .
( [ asd ] ) as well as the fact that the bare propagator does not depend on the renormalization scale yields = -_((),a_r ( ) ) , [ ren - eq ] where @xmath78 depends on @xmath86 explicitly as well as through @xmath73 and @xmath87 .
as will be shown later , the explicit @xmath86 dependence enters through the combination @xmath88 and the renormalized propagator may be written : s_r(p ; ) = s_r ( ; ; a_r ) .
[ sr - sr ] in partial differential form the renormalization group equation , eq .
( [ ren - eq ] ) , may then be written as : ( + ( ) - _ a ( ) + _ ( , a_r ) ) s_r(p ; ) = 0 .
[ ren - eq - part - mu ] equivalently , we may use eq .
( [ sr - sr ] ) in order to trade the partial derivative with respect to @xmath86 in eq .
( [ ren - eq - part - mu ] ) for a derivative with respect to momentum : ( 2 - ( ) + _ a ( ) - _ ( , a_r ) ) s_r ( ; ; a_r ) = 0 .
[ ren - eq - part - p ] we shall come back to the solution of this equation in a later section .
one - loop diagrams : a ) photon self - energy and b ) fermion self - energy . ]
having set the basic notations and goals , we summarize here the one - loop results obtained in ref . [ ] . as will be seen below
, all diagrams can be compactly expressed as a function of the one - loop massless propagator diagram defined as : = g ( , ) , [ 1loopmassless ] where @xmath73 and @xmath89 are arbitrary indices labeling the diagram , the power of @xmath90 follows from dimensional analysis and @xmath91 is the ( dimensionless ) coefficient function associated with the diagram .
the latter is known exactly in the one - loop case : g ( , ) = a ( ) = .
[ 1loopg ] the one - loop diagrams , see fig .
[ fig : one - loop ] , are defined by : [ 1loopdiagrams ] & & i_1^(q ) = - , [ pimunu-1l ] + & & -i_1(p ) = ( -ie^ ) ( -ie^ ) ( g _ - ) , [ sigma-1l ] with the above conventions and identities in hand , eqs . ( [ 1loopdiagrams ] ) can be straightforwardly computed in an arbitrary gauge .
the final result reads : [ 1loopdiagrams - s ] & & _
1(q^2 ) = - dg(1,1 ) , [ pimunu-1l - s ] + & & _
v1(p^2 ) = - g(1,1-_e ) .
[ sigma-1l - s ] as can be seen from eqs .
( [ 1loopdiagrams - s ] ) , the one - loop results are expressed in terms of two master integrals : @xmath100 for the photon self - energy and @xmath101 for the fermion self - energy where @xmath102 is given by eq .
( [ 1loopg ] ) .
these equations have been discussed in ref . [ ] and we refer the interested reader to this reference for more details .
these equations will be used in the next sections concerning the computation of the two - loop fermion self - energy and renormalization .
two - loop massless propagator diagram . ] in the one - loop case , after working out the numerator algebra , the considered diagrams could be expressed in terms of two simple scalar integrals @xmath100 and @xmath101 , see eqs .
( [ 1loopdiagrams - s ] ) .
these one - loop integrals are known exactly from eq .
( [ 1loopg ] ) . at the two - loop level
the situation is more complicated .
the two - loop massless propagator diagram is actually one of the basic building blocks of multi - loop calculations and has a long history , see the review ref .
[ ] . for arbitrary indices @xmath103 and
external momentum @xmath104 in a minkowski space - time of dimensionality @xmath11 the two - loop massless propagator diagram , see fig .
[ fig : j ] , reads : = - g(_1,_2,_3,_4,_5 ) , [ def:2loopmassless ] where , on the right - hand side , the power of @xmath104 follows from dimensionality and the function @xmath105 is the 2-loop ( dimensionless ) coefficient function of the diagram .
one of the main goals of multi - loop calculations is an exact evaluation of @xmath105 . when all indices are integers this function can be computed exactly using ibp identities .
@xcite the diagram can then be expressed in terms of recursively one - loop integrals as can be seen from the following well - known example : was actually first computed exactly in dimensional regularization with the help of the @xmath106-space gegenbauer polynomial technique in ref .
subsequently , it was computed exactly with the help of ibp in refs .
g(1,1,1,1,1 ) = c_d = , [ g11111 ] where @xmath107 denotes the coefficient function of the diagram . for arbitrary indices
its evaluation is however highly nontrivial : it can be represented @xcite as a combination of twofold series . in some simpler cases
explicit analytical results can be obtained .
@xcite as an example , the diagram @xmath108 can be expressed in a compact way for a special value of the non - integer index on the central line : @xcite g(1,1,1,1,_e ) = c_d= 3 , ( _ e = d_e/2 - 1 ) [ ilambda ] where @xmath109 is the trigamma function and @xmath110 is the digamma function .
the result of eq .
( [ ilambda ] ) elegantly covers the cases of rqed@xmath15 s where @xmath111 and in particular rqed@xmath8 where the gauge field does not renormalize . in the general case of rqed@xmath7 such a diagram appears in the calculation of the two - loop photon self - energy but with a more general index : @xmath112 where @xmath69 can be non - zero . in this case , the result is more complicated and reads : @xcite g(1,1,1,1 , ) & = & c_d= -2 ( _ e)(_e- ) ( 1 - 2_e+ ) [ kotikovg ] + & & , where the one - fold series corresponds to a generalized hypergeometric function , @xmath113 , of argument @xmath53 . the diagram evaluated in eq .
( [ kotikovg ] ) is one of the simplest among a class of complicated diagrams that have been considered in ref .
[ ] on the basis of a new development of the gegenbauer polynomial technique .
this class includes diagrams with two adjacent lines having integer indices while three other lines have arbitrary indices ; the corresponding coefficient function is given by : @xmath114 . as proved in ref .
[ ] all of these diagrams may be expressed in terms of generalized hypergeometric functions @xmath113 of argument @xmath53 .
for this class of diagrams , similar results have been found in ref . [ ] using an ansatz to solve the recurrence relations for the two - loop diagram .
however , beyond the simplest case with a single non - integer index represented by eq .
( [ kotikovg ] ) , explicit expressions for other diagrams of this class are unknown . as will be seen below , it turns out that the computation of the two - loop fermion self - energy for a general rqed@xmath7 requires the knowledge of a two - loop massless propagator diagram with two non - integer indices on non - adjacent lines : g(,1,,1,1 ) = c_d .
[ complicatedg ] using the results of ref .
[ ] an exact explicit expression for this diagram can be derived , see app .
[ app : g(alpha , beta ) ] for general formulas . as can be seen from the latter ,
the result is of higher complexity than the one of eq .
( [ kotikovg ] ) . for the sake of simplicity
we shall not consider the general solution here . instead , and in the spirit of ref .
[ ] , we shall focus on the less general but important case of rqed@xmath15 s for which @xmath111 .
as will be shown below , in this case a simpler solution can be found using integration by parts .
2-loop fermion self - energy diagrams : a ) bubble diagram , b ) rainbow diagram and c ) crossed photon diagram . ]
having set all the perturbative framework in sec .
[ sec : rqed ] and presented the two - loop massless propagator diagram in sec .
[ subsec : twoloop ] , we now proceed on computing the two - loop fermion self - energy . the latter consists of three diagrams that are displayed on fig . [ fig : rqed-2loop - self - energy ] . the first diagram , fig .
[ fig : rqed-2loop - self - energy]a , is the so - called bubble diagram which is defined as : & & -i_2a(p ) = [ sigma - a ] + & & ( -ie^ ) ( -ie^ ) ( g _ - ) i_1^(k ) ( g _ - ) , where the lower index @xmath115 in @xmath116 refers to 2-loop and @xmath117 is a one - loop photon self - energy insertion corresponding to eq .
( [ pimunu-1l - s ] ) .
this diagram is obviously gauge invariant .
this follows from current conservation which implies that @xmath118 so that all terms proportional to the gauge fixing term vanish .
using eq .
( [ sigmav ] ) and performing all the numerator algebra with the help of eqs .
( [ num - algebra ] ) yields : _
v2a(p^2 ) = dg(1,1)g(1,_-_e ) , [ sigma - a - s ] where the factor of @xmath96 is due to the fermion loop and the master integral , @xmath119 , is a simple product of two one - loop integrals which can be straightforwardly evaluated with the help of eq .
( [ 1loopg ] ) . the second diagram , fig .
[ fig : rqed-2loop - self - energy]b , is the so - called rainbow diagram which is defined as : -i_2b(p ) = ( -ie^ ) ( -i_1(p+k ) ) ( -ie ^ ) ( g _ - ) , [ sigma - b ] where @xmath120 is a one - loop fermion self - energy insertion corresponding to eq . ( [ sigma-1l - s ] ) .
this diagram is not gauge invariant .
proceeding along the same lines as for the bubble diagram yields : _
v2b(p^2 ) & = & ( - ) ^2g(1,1-_e)g(1-_e , _ ) , [ sigma - b - s ] where the master integral , @xmath121 , is also a simple product of two one - loop integrals which can be straightforwardly evaluated with the help of eq .
( [ 1loopg ] ) . finally , the third diagram , fig .
[ fig : rqed-2loop - self - energy]c , which is the so - called crossed photon diagram is a truly two - loop diagram . it is defined as : -i_2c(p ) = & & ( -ie^ ) ( g _ - ) [ sigma - c ] + & & ( -ie^ ) ( g _ - ) ( -ie^)(-ie^ ) .
proceeding along the same lines as for the two previous diagrams yields : _
v2c(p^2 ) & = & - [ sigma - c - s ] + & & \ { ( d_e-4 - 4 -2+^2 )
g^2(1,1-_e ) + . + & & .
+ ( d_e + 4 + 8 + + 4 -(d_-3)^2 ) g(1,1-_e)g(1-_e , _ ) + .
+ 4g(1-_e,1,-1-_e,1,1 ) - 4 g(1-_e,1,-_e,1,1)+(8-d_e)g(-_e,1,-_e,1,1 ) } , where , besides the primitively one - loop master integrals , @xmath122 and @xmath121 , complicated two - loop propagator diagrams of the type of eq .
( [ complicatedg ] ) enter the expression of the self - energy .
it is interesting to notice that none of the complicated terms depend on the gauge parameter , @xmath123 .
that is , working in the feynman gauge , @xmath124 , simplifies the calculations but does not reduce the complexity of the diagrams . in order to further reduce the complicated two - loop diagrams appearing in eq .
( [ sigma - c - s ] ) we use an ibp identity which follows from the homogeneity of the two - loop propagator diagram , eq .
( [ def:2loopmassless ] ) , in @xmath104 , see ref .
[ ] and ref .
[ ] for a review . in graphical form
, this ibp reads : ( + _ 4-_1-_2-_5 ) & & = _ 3 ( - ) [ ibp:2loop - homog ] + & & + & & + ( - _ i=1 ^ 5_i ) ( -p^2)^-1 ( - ) , + & & where @xmath125 on the right - hand side of the equation denotes the increase or decrease of a line index by @xmath53 with respect to its value on the left - hand side .
this expression slightly simplifies in the cases we are interested in : ( 1 - _ e - _ - ) & & = ( - ) [ ibp:2loop - b ] + & & + & & + ( 3- 3_e - 3_- - ) ( -p^2)^-1 ( - ) , + & & where we have used eq .
( [ d - params ] ) and the fact that , in @xmath104-space , a line with zero index shrinks to a point . as a consequence , on the right - hand side of eq .
( [ ibp:2loop - b ] ) all diagrams except the third one are recursively one - loop diagram . in order to see how eq .
( [ ibp:2loop - b ] ) can be usefully implemented , let s consider the case where @xmath126 . in this case
( [ ibp:2loop - b ] ) reduces to : - _ & = & ( 1-_e ) ( - ) [ ibp:2loop - c ] + & & + & & + ( 1 - _ e - 3 _ ) ( -p^2)^-1 ( - ) , + & & where we have used the symmetry of the two - loop diagram : @xmath127 .
we then see that the truly two - loop diagram appearing on the right - hand side of eq .
( [ ibp:2loop - c ] ) has a coefficient function @xmath128 .
it is therefore one of the complicated diagrams appearing in eq .
( [ sigma - c - s ] ) that we would like to compute . at this point , it is important to notice that the function @xmath128 is associated with an ultra - violet singular two - loop diagram in the limit @xmath111 . indeed , power counting shows that this diagram is proportional to : @xmath129 , _
i.e. _ , diverges logarithmically . on the other - hand , the two - loop diagram appearing on the left - hand side of eq .
( [ ibp:2loop - c ] ) , the coefficient function of which is given by : @xmath130 , is proportional to @xmath131 ; it is therefore ultra - violet convergent in the limit @xmath111 . as a consequence , the laurent series associated to its @xmath31-expansion reduces to the regular part : g(1-_e,1,1-_e,1,1 ) = _ n=0^c_n(_e)_^n . for @xmath132 , which applies to qed@xmath18 and qed@xmath19
, this diagram corresponds to the well - known eq .
( [ g11111 ] ) . on the other hand , for an arbitrary non - integer @xmath46 , for example @xmath49 in the case of rqed@xmath8
, it belongs to the class of complicated diagrams , see eq .
( [ complicatedg ] ) and related discussion around this equation , whose explicit analytic expression , and hence the coefficients @xmath133 , is presently unknown .
however , the fact that this diagram is convergent together with its coefficient @xmath31 , implies that the left - hand side of eq .
( [ ibp:2loop - c ] ) vanishes in the limit @xmath111 .
hence , from the right - hand side of eq .
( [ ibp:2loop - c ] ) we find that : & = & - ( - ) + & & + & & - .
[ ibp:2loop - d ] + & & after some algebra , the recursively one - loop diagrams appearing in the left - hand side of eq .
( [ ibp:2loop - d ] ) can be expressed in terms of the one - loop master integrals entering the expression of the fermion self - energy eq .
( [ sigma - c - s ] ) .
the result reads : g(1-_e,1,-_e,1,1 ) & = & -2g(1,1-_e)g(1-_e , _ ) + g^2(1,1-_e ) [ g2 ] + & + & g(1-_e,1,1-_e,1,1 ) , where the last term does not contribute in the limit @xmath17 .
hence , for rqed@xmath15 and in particular in the case of rqed@xmath8 where the last diagram in eq .
( [ g2 ] ) is non - trivial , the function @xmath128 is known for arbitrary @xmath46 to @xmath134 .
the knowledge of these lowest order terms is enough , _
e.g. _ , to compute the anomalous scaling dimension of the fermion field at two - loop ( see below for more ) .
we may proceed in a similar way for the two other complicated diagrams appearing in eq .
( [ sigma - c - s ] ) .
the corresponding coefficient functions , @xmath135 and @xmath136 , are proportional to @xmath137 and are strongly uv divergent .
they can both be expressed in terms of @xmath128 . in order to see this
, we consider the case where @xmath138 and @xmath139 for which eq .
( [ ibp:2loop - b ] ) reduces to : ( 1- _ ) & = & ( 1-_e ) ( - ) [ ibp:2loop - e ] + & & + & & + ( 2 - _ e - 3 _ ) ( -p^2)^-1 ( - ) .
+ & & eq . ( [ ibp:2loop - e ] ) shows that there is a simple relation , _
i.e. _ , involving recursively one - loop diagrams , between two of the complicated diagrams , @xmath135 and @xmath128 , entering the fermion self - energy , eq .
( [ sigma - c - s ] ) . expressing the , still unknown , coefficient function @xmath135 in terms of @xmath128 that we have computed above ,
( [ g2 ] ) , yields , in graphical form : & = & - ( - ) + & & + & & + . [ ibp:2loop - f ] after some algebra , the recursively one - loop diagrams appearing in the right - hand side of eq .
( [ ibp:2loop - f ] ) can also be expressed in terms of the one - loop master integrals that enter the expression of the fermion self - energy eq .
( [ sigma - c - s ] ) . together with eq .
( [ g2 ] ) , the final result reads : g(1-_e,1,-1-_e,1,1 ) & = & - ( + ) g(1,1-_e)g(1-_e , _ ) [ g1 ] + & + & ( + ) g^2(1,1-_e ) + & + & g(1-_e,1,1-_e,1,1 ) . finally , let s consider the case where @xmath140 and @xmath141 for which eq .
( [ ibp:2loop - b ] ) reduces to : - _ & = & -_e ( - ) [ ibp:2loop - g ] + & & + & & + ( 2 - _ e - 3 _ ) ( -p^2)^-1 ( - ) .
+ & & from eq .
( [ ibp:2loop - g ] ) we now have a relation between @xmath136 and @xmath128 . expressing the former as a function of the later , yields , in graphical form : & = & + ( - ) + & & + & & - . [ ibp:2loop - h ] similarly to the previous cases , the recursively one - loop diagrams appearing in the right - hand side of eq .
( [ ibp:2loop - h ] ) can be expressed in terms of the one - loop master integrals entering the expression of the fermion self - energy eq .
( [ sigma - c - s ] ) . together with eq .
( [ g2 ] ) , the final result reads : g(-_e,1,-_e,1,1 ) & = & -g(1,1-_e)g(1-_e , _ ) [ g3 ] + & + & ( + d_- d_e ) g^2(1,1-_e ) + & + & g(1-_e,1,1-_e,1,1 ) . eqs .
( [ g2 ] ) , ( [ g1 ] ) and ( [ g3 ] ) constitute the central results of this paragraph .
they show that all three complicated diagrams appearing in eq .
( [ sigma - c - s ] ) are related : they can all be expressed in terms of the ultra - violet convergent diagram @xmath130 with a factor proportional to @xmath31 . with the help of eqs .
( [ g2 ] ) , ( [ g1 ] ) and ( [ g3 ] ) the two - loop self - energy associated with the crossed photon diagram , eq .
( [ sigma - c - s ] ) , reads : & & _
v2c(p^2 ) = - \ { g^2(1,1-_e ) + & & .
+ g(1,1-_e)g(1-_e , _ ) . + & & . - g(1-_e,1,1-_e,1,1 ) } .
the sum of all the self - energy contributions , eqs .
( [ sigma - a - s ] ) , ( [ sigma - b - s ] ) and eq .
( [ sigma - c - s2 ] ) , yields the total two - loop fermion self - energy in an arbitrary gauge : _
v2(p^2 ) & = & \ { dg(1,1)g(1,_-_e ) .
[ sigma-2l - s ] + + & & + g^2(1,1-_e ) + & & .
+ g(1,1-_e)g(1-_e , _ ) + .
+ g(1-_e,1,1-_e,1,1 ) } . as explained in the previous paragraph , the last line of eq .
( [ sigma-2l - s ] ) contains the uv - convergent diagram @xmath130 with a prefactor proportional to @xmath142 .
it therefore vanishes for all rqed@xmath15 and in particular in rqed@xmath8 which enables us to avoid computing the complicated diagram @xmath143 and obtain an expansion valid to @xmath134 for the two - loop fermion self - energy .
( [ sigma-2l - s ] ) simplifies for usual qeds where @xmath144 corresponding to @xmath132 . in this case
, the complicated diagram reduces to the well known @xmath145 which is given by eq .
( [ g11111 ] ) and eq .
( [ sigma-2l - s ] ) becomes : _
v2(p^2 ) = ( d_-2 ) \ { & & 2n_fg(1,1)g(1 , _ ) -(d_-6+(d_-2)a^2)g^2(1,1 ) .
[ sigma-2l - s - qedd ] + & & + .
( a^2(3d_-8 ) - d_-4)g(1,1)g(1 , _ ) } , where eq .
( [ nf ] ) has been used .
( [ sigma-2l - s - qedd ] ) agrees with well known results in the literature , see _
e.g. _ , ref .
we now compute the renormalization constant associated to the fermion field @xmath70 , the related anomalous dimension @xmath146 and renormalized propagator @xmath147 up to two loops . to achieve this task we use for rqed well known methods used for usual qeds , see , _
e.g. _ , ref .
[ ] . in the following , we shall expand the one - loop and two - loop fermion self - energies , for fixed @xmath46 , in laurent series in @xmath31 .
for this purpose , we introduce two functions , @xmath148 and @xmath149 , which are defined with the help of the expressions of the one - loop ( eq . ( [ sigma-1l - s ] ) ) and two - loop ( eq . ( [ sigma-2l - s ] ) ) fermion self - energies as follows : [ ren - sigmas ] _
v1(p^2 ) & = & _ 1(_e,_,a ) , [ ren : def : sigma1 ] + _ v2(p^2 ) & = & _ 2(_e,_,a ) .
[ ren : def : sigma2 ] as will be seen shortly , it is important to single out in the function @xmath148 the part which depends on the ( bare ) gauge fixing parameter @xmath24 . from eq .
( [ sigma-1l - s ] ) , this yields [ ren - sigma1s ] & & _
1(_e,_,a ) = _ 1(_e , _ ) + a _ 1(_e , _ ) , [ sigma1-sep ] + & & _ 1(_e , _ ) = ( 1-_e)g(1,1-_e ) , [ ren : sigma1 ] + & & _ 1(_e , _ ) = -(1-_e)(1-_e-_)g(1,1-_e ) .
on the other hand , in the 2-loop function @xmath149 we may single out the three contributions corresponding to the three diagrams of fig .
[ fig : rqed-2loop - self - energy ] whose expressions are given by eqs .
( [ sigma - a - s ] ) , ( [ sigma - b - s ] ) and ( [ sigma - c - s2 ] ) .
this yields : [ ren - sigma2s ] & & _
2(_e,_,a ) = ^(2)_a(_e,_)+^(2)_b(_e,_,a)+^(2)_c(_e,_,a ) , + & & ^(2)_a(_e , _ ) = -4n_f^2(1-_e)g(1,1)g(1,_-_e ) , + & & ^(2)_b(_e,_,a ) = -^2(1-_e)g(1,1-_e)g(1-_e , _ ) , + & & ^(2)_c(_e,_,a ) = ^2(1-_e ) ( 1-_e- _ ) + & & \ { g^2(1,1-_e ) + & & - g(1,1-_e)g(1-_e , _ ) + & & .- g(1-_e,1,1-_e,1,1 ) } . these equations are no more than a re - writing of the two - loop self - energies , eqs .
( [ sigma - a - s ] ) , ( [ sigma - b - s ] ) and ( [ sigma - c - s2 ] ) , in terms of the @xmath150-parameters instead of the dimensions . in order to implement the renormalization procedure all bare quantities ( charge and gauge fixing parameter ) appearing in the self - energies eqs .
( [ ren - sigmas ] ) have to be expressed in terms of renormalized ones : = z_(())^2_e^_e _ , a = z_a ( ( ) ) a_r ( ) , where eqs .
( [ ren_coupling ] ) and ( [ ren_constants ] ) were used . for an expansion to @xmath151 it is enough to know the coupling and gauge field renormalizations at one - loop . from ref .
[ ] they read : z _ ( ) = z_a^-1 ( ) = 1+z _ + ( ^2 ) , _ a ( ) = 2z _ , ( ) = -2_+ 2z _ , [ 1loopren - gauge ] where the coefficient @xmath152 depends of the theory under consideration : z_= ( _ 4),z_= 0(_4,3 ) .
[ 1loopren - gauge - zal ] summing the one - loop and two - loop contributions , the total self - energy up to two loops can then be written as : _
v(p^2 ) = e^(_e - l_p)__1(a_r ) + ( ) ^2 ( e^(_e - l_p)__1+e^2(_e - l_p)__2(a_r ) ) + ( ^3 ) , [ self - energy - exp ] where @xmath153 and the functions @xmath154 and @xmath149 now depend on the _ renormalized _ gauge fixing parameter . with the help of eq . ( [ self - energy - exp ] ) the expansion of the fermion propagator eq .
( [ fermion - prop+sigma ] ) reads : -i s(p ) = 1 + e^(_e - l_p)__1(a_r ) + ( ) ^2 + ( ^3 ) .
[ fermion - prop - exp1 ] the laurent series associated with the various functions entering eq .
( [ fermion - prop - exp1 ] ) can be written as : [ exps ] & & e^(_e - l_p)__1 = + c_11 + c_12_+ , e^(_e - l_p)__1
= + c_11
+ c_12_+ , [ c1j ] + & & e^2(_e - l_p ) _ ( _ 1 ^ 2 + _ 2 ) = + + c_22 + , [ c2j ] where the dots indicate higher order terms in @xmath69 . in eq .
( [ c2j ] ) the two - loop contribution has been expanded up to @xmath155 in @xmath69 which is all we need for two - loop renormalization and is within the accuracy of our loop calculations in rqed@xmath15 . with these notations eq .
( [ fermion - prop - exp1 ] ) can then be re - written as : -i s(p ) = 1 + ( + c_11 + c_12_+ ) + ( ) ^2 ( + + c_22 + z_c_12 + ) + ( ^3 ) , [ def : series : s ] where the one - loop coefficients @xmath156 are defined by eqs .
( [ c1j ] ) and the coefficients @xmath157 , which also depend on @xmath87 , are defined by eq .
( [ c2j ] ) . in the @xmath68 scheme
all singular terms are factorized in the renormalization constant @xmath70 while the regular terms contribute to the renormalized propagator @xmath78 , yielding general expressions of the form : [ ren - z+s ] z_(,a_r ) & = & 1 + + ( ) ^2 ( + ) + ( ^3 ) , [ def : series : z ] + -i s_r(p ; ) & = & 1 + ( r_1 + r_11_+ ) + ( ) ^2 ( r_2 + ) + ( ^3 ) , [ def : series : sr ] where the coefficients @xmath158 and @xmath159 are to be determined .
this can be done with the help of eq .
( [ mpropr ] ) by identifying , at a given loop order , the coefficients of equal powers of @xmath31 .
the result reads : [ ren - z+r ] & & z_1 = c_10 , z_20 = c_20 + z_c_10 , z_21 = c_21 + z_c_11- c_10 c_11 , [ def : zcoeff ] + & & r_1 = c_11 , r_11 = c_12 , r_2 = c_22+z_c_12- c_10 c_12 .
[ def : rcoeff ] we are now in a position to present a general expression for the anomalous scaling dimension of the fermion field . in partial derivative form
the latter reads _ ( ( ) , a_r ( ) ) = = ( ( ) ) - _ a ( ) , [ gamma_psi_2l ] where the one - loop beta - function and anomalous scaling dimension of the gauge field are given by eqs .
( [ 1loopren - gauge ] ) . from eqs .
( [ def : series : z ] ) , ( [ def : zcoeff ] ) and ( [ 1loopren - gauge ] ) , we recover the fact , well known in usual qed , that the simple poles of @xmath70 determine the anomalous scaling dimension : _ ( , a_r ) = -2z_1 - 4z_21 ( ) ^2 + ( ^3 ) , [ gamma_psi - z ] and that , together with the beta function , they also determine the higher order poles : z_20 = .
[ z - constraint ] the constraint eq .
( [ z - constraint ] ) comes from the fact that the anomalous dimension has to be finite as @xmath160 . in this paragraph
we shall compute the anomalous dimension of the fermion field and renormalized fermion propagator in a general theory of rqed@xmath15 . in order to do so , we expand the @xmath161 functions of eqs .
( [ ren - sigma1s ] ) and ( [ ren - sigma2s ] ) in laurent series in @xmath69 for arbitrary @xmath67 .
the coefficients of the expansions read : [ ren - sigmas - rqed4de ] & & e^(_e - l_p)__1 = - ( |l_p - ) , [ sigma-rqed4de ] + & & + ( |l_p ( |l_p - ) + 2 _ 2 -3_2(2-_e ) + ) _ + ( _ ^2 ) , + & & e^(_e - l_p)__1
= -+|l_p - ( |l_p^2 + 2_2 - 3 _ 2(2-_e ) ) _ + ( _ ^2 ) , [ sigma-rqed4de ] + & & e^2(_e - l_p ) _ ( _ 1 ^ 2 + _ 2 ) = ( a_r - ) ^2 - ( ( a_r - ) ^2|l_p + 2a_r .
[ sigma2-rqed4de ] + & & .
- 2 + 2 n_f k_1 ) + & & + ( a_r-)^2 ( |l_p^2+_2- 2_2(2-_e ) ) + 4a_r ( |l_p - ) + & & + 4(_2(1-_e ) - _ 2 ) -4|l_p + 2 + & & -2 + 4n_f k_1 ( |l_p+(_1 - _ 2 ) + - ) + ( _ ) , where [ ren - params - rqed4de ] & & |l_p = l_p - _ 1(2-_e)+ _ 1(1 ) , k_1 = , [ ren - params - rqed4de - lp+k1 ] + & & _
1=_1(1-_e)-_1(1 ) , _ 2=_1(_e)-2_1(1-_e)+2_1(2 - 2_e ) -_1(1 ) , [ ren - params - rqed4de - psis ] and @xmath162 and @xmath163 are the digamma and trigamma functions , respectively . from eqs .
( [ ren - sigmas - rqed4de ] ) the coefficients @xmath164 and @xmath157 can be deduced . focusing for the moment on the singular contributions , which are obtained by combining eqs .
( [ ren - sigmas - rqed4de ] ) with eqs .
( [ def : zcoeff ] ) , yields : [ ren - z - rqed4de ] & & z_1 = -(a_r - ) , z_20 = ( a_r - ) ^2 + z _ , [ z1+z20-rqed4de ] + & & z_21 = -2n_f k_1 + 2 -z_(|l_p - ) , [ z21-rqed4de ] where all gauge depend terms cancel out from @xmath165 . in order for the consistency relation , eq .
( [ z - constraint ] ) , to be valid we see from eqs .
( [ z1+z20-rqed4de ] ) that we must either have @xmath166 if @xmath167 which is the case of qed@xmath18 or @xmath168 if @xmath169 which is the case of rqed@xmath8 , see eq . (
[ 1loopren - gauge - zal ] ) . taking this constraint into account
we see that momentum dependent terms also cancel out from @xmath165
. then , the general expression of the anomalous dimension of the fermion field in rqed@xmath15 reads _
( , a_r ) = 2(a_r - ) + 8 ( n_f k_1 - ) ( ) ^2 + ( ^3 ) .
[ asd - rqed4de ] it is important that eq .
( [ asd - rqed4de ] ) does not depend on the external momenta and digamma functions .
this property is an extension of the rule @xcite for the results of the anomalous dimensions in standard qft ( @xmath170 and @xmath171 ) , where the corresponding anomalous dimensions do not depend on the external momenta and ( in @xmath172-scheme ) on @xmath74 and @xmath173 .
note that in the framework of rqed@xmath8 there is a contribution of @xmath173 in the corresponding anomalous dimension ( see eq .
( [ asd - rqed43 ] ) below ) but this contribution appears , not from the expansion of @xmath174-functions in @xmath69 but from the factor @xmath175 in the expression @xmath100 , see app .
[ app - gfuncs ] .
finally , the finite part can be obtained in a similar way by combining eqs .
( [ ren - sigmas - rqed4de ] ) with eqs .
( [ def : rcoeff ] ) .
this yields : [ ren - r - rqed4de ] r_1 & = & ( a_r-)|l_p + , [ r1-rqed4de ] + r_11 & = & -(a_r-)(|l_p^2 + 2_2 - 3_2(2-_e ) ) - ( |l_p- ) , [ r11-rqed4de ] + r_2 & = & ( a_r-)^2 ( |l_p^2+_2-_2(2-_e ) ) + 2a_r|l_p + 4(_2(1-_e ) - _ 2 ) + & - & 2|l_p -2 - 2 + & + & 4n_f k_1 ( |l_p+(_1 - _ 2 ) + - ) . [ r2-rqed4de ]
a few remarks are in order concerning the above results starting from eqs .
( [ ren - sigmas - rqed4de ] ) . the laurent series in @xmath69
are plagued by singularities for some values of @xmath67 .
the singularity at @xmath176 ( case of a 0-brane or quantum mechanics ) reflects the fact that the self - energy of a point - like particle is ill - defined already at one - loop .
the singularity at @xmath177 ( case of a 1-brane or rqed@xmath38 ) appears starting from two - loop and is related to the zero - width of the filament , _
i.e. _ , the higher powers of @xmath178 , coming from the effective free gauge field propagator ( [ gauge - field - prop0 ] ) , as the order increases .
finally , at two - loop , a singularity appears for @xmath170 ( case of qed@xmath18 ) as can be seen in particular from the last term in eq .
( [ r2-rqed4de ] ) .
this singularity is gauge invariant and can be traced back to the uv behaviour of the one - loop polarization operator entering the bubble diagram which is contained in the master integral @xmath100 , see app .
[ app - gfuncs ] .
the above results are therefore valid only in application to rqed@xmath8 . in the case of massless rqed@xmath8 ( @xmath49 and @xmath111 ) the @xmath69-expansions of eqs .
( [ ren - sigmas - rqed4de ] ) read : [ ren - sigmas - rqed43 ] e^(_e - l_p)__1 & = & + + ( |l_p ( |l_p - ) - 7_2 + ) _ + ( _ ^2 ) , [ sigma-rqed43 ] + e^(_e - l_p)__1 & = & -+|l_p - _ + ( _ ^2 ) , [ sigma-rqed4 ] + e^2(_e - l_p ) _ ( _ 1 ^ 2 + _ 2 ) & = & - + + & + & ( |l_p^2 -_2 + 8 ) + ( |l_p - ) + _ 2 + |l_p + & + & 8 _ 2 n_f ( |l_p + 2 - 4 ) - + ( _ ) , [ sigma2-rqed43 ] where @xmath179 . the coefficients then read : [ ren - z+r - rqed43 ] & & z_1 = , z_20 = , z_21 = -4_2n_f - , [ def : zcoeff - rqed43 ] + & & r_1 = -|l_p , r_11 = |l_p^2 -|l_p-7_2 - 6a_r + , + & & r_2 = ( |l_p^2 - 2_2 + 4 ) + 4 -8_2n_f(|l_p+2 - 4 ) - .
[ def : rcoeff - rqed43 ] the corresponding renormalization constant and renormalized fermion propagator read : [ 2loopren - rqed43 ] & & z_(,a_r ) = 1 + + ( ) ^2 + ( ^3 ) , [ 2loopren - rqed43-zpsi ] + & & -i s_r(p ; ) = 1 + [ 2loopren - rqed43-sr ] + & & + ( ) ^2 + ( ^3 ) .
finally , the anomalous scaling dimension of the fermion field in rqed@xmath8 reads : _ ( , a_r ) = 2 + 16(_2 n_f + ) ( ) ^2 + ( ^3 ) .
[ asd - rqed43 ] as already noticed in the general formulas , this anomalous dimension does not depend on the external momentum and the two - loop contribution does not depend on the choice of gauge .
in order to see how it affects the momentum dependence of the fermion propagator we solve eq .
( [ ren - eq - part - p ] ) . for this purpose , we use the fact that : @xmath180 , as the coupling and gauge field do not renormalize in rqed@xmath8 .
the solution then reads : s_r(p ; ) = s_r(1 ; ; a ) ( ) ^_(,a ) , [ ren - prop - rqed43 ] where @xmath146 is given by eq .
( [ asd - rqed43 ] ) .
the case of massless qed@xmath18 corresponds to : @xmath132 and @xmath111 .
starting from eq .
( [ sigma-2l - s - qedd ] ) the @xmath69-expansions of the self - energies read : [ ren - sigmas - qed4 ] e^(_e - l_p)__1 & = & 0 , [ sigma-qed4 ] + e^(_e - l_p)__1 & = & -+l_p-1+_+ ( _ ^2 ) , [ sigma-qed4 ] + e^2(_e - l_p ) _ ( _ 1 ^ 2 + _ 2 ) & = & + + + & + & a_r^2 ( l_p(l_p-2 ) - _ 2 + 3 ) -2l_p(n_f+ ) + n_f+ + ( _ ) , [ sigma2-qed4 ] from which the coefficients @xmath164 and @xmath157 can be deduced ; in particular , the coefficients @xmath181 in qed@xmath18 . substituting these coefficients in eqs
( [ ren - z+r ] ) , yields : [ ren - z+r - qed4 ] & & z_1 = -a_r , z_20 = , z_21 = n_f + , [ def : zcoeff - qed4 ] + & & r_1 = a_r(l_p-1 ) , r_11 = ( _ 2 - l_p(l_p-2)-4 ) , + & & r_2 = a_r^2(1+l_p(l_p-2 ) ) -2l_p(n_f+ ) + n_f + . [
def : rcoeff - qed4 ] from eq .
( [ def : zcoeff - qed4 ] ) we see that eq .
( [ z - constraint ] ) is indeed satisfied .
moreover , similarly to the case of qed@xmath8 , while @xmath182 is gauge - variant , the gauge dependence cancels out in @xmath165 . substituting the coefficients of eqs .
( [ ren - z+r - qed4 ] ) in eqs .
( [ ren - z+s ] ) yields the following renormalization constant and renormalized fermion propagator : [ 2loopren - qed4 ] z_(,a_r ) & = & 1 -a_r + ( ) ^2 + ( ^3 ) , [ 2loopren - qed4-zpsi ] + -i s_r(p ; ) & = & 1 + + + & & + ( ) ^2 + ( ^3 ) , [ 2loopren - qed4-sr ] finally , from eqs .
( [ def : zcoeff - qed4 ] ) and ( [ gamma_psi - z ] ) the anomalous scaling dimension of the fermion field in qed@xmath18 is recovered : _ ( , a_r ) = 2a_r - 4(n_f + ) ( ) ^2 + ( ^3 ) .
[ asd - qed4 ] as in the case of rqed@xmath8 , see eq .
( [ asd - rqed43 ] ) , the fermion anomalous dimension of qed@xmath18 does not depend on the external momentum and the two - loop contribution does not depend on the choice of gauge .
furthermore , eq . ( [ asd - qed4 ] ) is transcendentally simpler than eq .
( [ asd - rqed43 ] ) as no @xmath173 appears in eq .
( [ asd - qed4 ] ) .
finally , we consider the case of massless qed@xmath19 .
such a model differs considerably from the previous ones because it is super - renormalizable .
nevertheless we shall proceed along the same lines as the previous , renormalizable , models . in qed@xmath19
we have @xmath170 and the expansion parameter is @xmath183 .
( [ ren - sigmas ] ) can then be written as : [ ren - sigmas - qed3 ] _
v1(p^2 ) & = & _ 1(_e,_,a ) = e^(_e - l_p)__1(_e,_,a ) , [ ren : def : sigma1-qed3 ] + _ v2(p^2 ) & = & _ 2(_e,_,a ) = ( ) ^2 e^2(_e - l_p)__2(_e,_,a ) .
[ ren : def : sigma2-qed3 ] where @xmath184 , @xmath185 is a momentum - dependent dimensionless coupling constant ( @xmath55 has dimension of mass in qed@xmath19 ) defined as : = , [ coupl - qed3 ] and we have used the fact that @xmath168 for qed@xmath19 , see eq . ( [ 1loopren - gauge ] ) . with the help of eq .
( [ sigma-2l - s - qedd ] ) , the expansions read : [ exp - sigmas - qed3 ] & & e^(_e - l_p)__1 = - + ( l_p - ( 4)+2 ) _ + ( _ ^2 ) , [ exp-1loop - qed3 ] + & & e^2(_e - l_p ) _ ( _ 1 ^ 2 + _ 2 ) = - + ( a^2 + ( 3l_p-2)n_f + - 7 ) + ( _ ) .
[ exp-2loop - qed3 ] from these results we see that a singularity appears only at two - loop : the @xmath186 pole in eq .
( [ exp-2loop - qed3 ] ) .
this singularity is gauge independent and has a coefficient proportional to @xmath34 . from eq .
( [ sigma-2l - s - qedd ] ) it can therefore be traced back to the bubble diagram which involves the master integrals @xmath100 and @xmath187 . while @xmath100 is finite in three - dimensional qed , the one - loop master integral @xmath187 is indeed divergent in the limit @xmath188 : g(1 , _ ) = . at this point
, the singularity looks like a uv one . indeed , coming back to the general expression for the one - loop master integral ( [ 1loopg ] ) and expressing it in terms of gamma functions yields : g ( , ) = , [ 1loopg+ ] where @xmath189 .
dimensional analysis then shows that a pole in the first gamma function in the numerator of ( [ 1loopg+ ] ) is associated with a uv singularity whereas a pole in either of the two other gamma functions in the numerator is associated with an ir singularity . for @xmath170 , @xmath190 and @xmath191
, we then see that the singularity is in the first gamma function of the numerator : g(1 , _ )
it turns out , however , that the present example is one in which there is an interchange between uv and ir types of singularities . indeed , in dimensional regularization
, both of these singularities correspond to poles of @xmath174-functions . to see the interchange ,
let s consider the example in more details .
the above considered uv type of singularity was actually related to our choice of master integral : we took @xmath192 in ( [ sigma-2l - s - qedd ] ) . in our calculations , however , it is the integral @xmath193 that is involved . on the one hand
, it can be related to the integral @xmath192 by using the simple property g(1,+1 ) = - g(1 , ) . on the other hand ,
the singularity in @xmath193 is an ir one . indeed : g(1,_+1 ) = = = - , and g(1,_+1 ) = .
hence , in qed@xmath19 the singularity in the two - loop fermion self - energy is of ir origin , as it was shown earlier in refs .
[ ] . this is to be contrasted with the cases of qed@xmath18 and rqed@xmath8 where the corresponding singularities are of the uv type . from equations ( [ exp - sigmas - qed3 ] ) we find the following coefficients : [ ren - z+r - qed3 ] & & z_1 = 0 , z_20 = 0 , z_21 = c_21 = - , [ zcoeff - qed3 ] + & & r_1 = - , r_11 = ( l_p - ( 4)+2 ) , r_2 = c_22= ( a^2 + ( 3l_p-2)n_f + - 7 ) .
[ rcoeff - qed3 ] this yields the following result for the renormalized fermion
propagator : -i s_r(p ; ) & = & 1 -(1 - ( l_p-(4)+2)_+ ) + ( a^2 + ( 3l_p-2)n_f + - 7 ) + ( ^3 ) + & = & 1 -(1 - ( l_p-(4)+2)_+ ) + ( a^2 + ( 3l_p-2)n_f + - 7 ) + ( e^6 ) . similarly , the renormalization constant reads : z_= 1 - ( ) ^2 + ( ^3 ) = 1 - ^2 + ( ^3 ) .
finally , the anomalous scaling dimension reads : _
= ( ) ^2 + ( ^3 ) = + ^2 + ( ^3 ) .
[ asd - qed3 ] interestingly , because there is no one - loop contribution , this anomalous dimension is fully gauge - invariant . in order to see how it affects the momentum dependence of the fermion propagator , we combine eq .
( [ sr - sr ] ) with the general solution of eq .
( [ ren - eq - part - p ] ) .
this yields : s_r(p ; ) = s_r(1;;a)e^ _ 0^ ( ) ( ) _ ( ( p;),a ) , [ ren - prop - qed3:def ] where we have taken into account the fact that , in qed@xmath19 , @xmath194 and the lack of running of the coupling constant @xmath55 at the considered level of accuracy .
the momentum dependence of this coupling constant implies that perturbation theory is valid , _
i.e. _ , @xmath195 , as long as : @xmath196 , _
i.e. _ , for large euclidean momenta . with the help of eq .
( [ asd - qed3 ] ) , the asymptotic form of the dressed fermion propagator defined in the left - hand side of ( [ ren - prop - qed3:def ] ) reads : s_r(p ; ) = s_r(1;;a)e^e^- .
[ ren - prop - qed3 ] deep in the uv , the momentum dependence of the dressed propagator is essentially the one of a free fermion in accordance with the fact that qed@xmath19 is asymptotically free .
the central result of this paper is the formula , eq .
( [ sigma-2l - s ] ) , for the two - loop fermion self - energy @xmath197 of massless rqed@xmath7 in an arbitrary gauge .
this formula was derived using simple ibp relations .
it allowed us , in the limit @xmath17 , to compute exactly @xmath197 of massless rqed@xmath15 without calculation of the complicated two - loop diagram @xmath143 ( the last term in eq .
( [ sigma-2l - s ] ) ) . from eq .
( [ sigma-2l - s ] ) , as well as the one - loop self - energy eq .
( [ sigma-1l - s ] ) , general expressions were derived for the fermion anomalous scaling dimension ( [ asd - rqed4de ] ) and the renormalized fermion propagator , eqs . ( [ ren - r - rqed4de ] ) and ( [ def : series : sr ] ) , in the limit @xmath17 .
these results were then applied to rqed@xmath8 and compared with the cases of ( renormalizable ) qed@xmath18 and ( super - renormalizable ) qed@xmath19 . in all cases ,
the two - loop contribution to the anomalous dimension was found to be gauge invariant .
the latter was also shown to be transcendentally more complex in rqed@xmath8 than in usual qeds as witnessed by the appearance of @xmath173 in eq .
( [ asd - rqed43 ] ) with respect to eqs .
( [ asd - qed4 ] ) and ( [ asd - qed3 ] ) . from the condensed matter physics point of view , as explained in the introduction , the massless rqed@xmath8 model describes the ultra - relativistic limit of undoped graphene .
our results are a first step towards a rigorous understanding of interaction corrections to the spectral properties of graphene in the ultra - relativistic limit . because of the lorentz invariance of the present model there is no renormalization of the fermi velocity
. has been done in the recent preprint by e. barnes , e. h. hwang , r. throckmorton and s. das sarma , arxiv:1401.7011 [ cond-mat.mes-hall ] . ]
the effect of interactions manifests at the level of the finite part of the fermion propagator as well as in the anomalous scaling dimension of the fermion field , eq .
( [ asd - rqed43 ] ) .
the later indicates how radiative corrections affect the momentum dependence of the dressed fermion propagator eq .
( [ ren - prop - rqed43 ] ) . for rqed@xmath8
our present results yield a gauge - variant dressed propagator coming from the one - loop contribution ; the two - loop contribution , on the other hand , is gauge - invariant and positive ( this is to be contrasted with the qed@xmath19 result which is fully gauge - invariant , see eq .
( [ ren - prop - qed3 ] ) ) . from the field theory point of view
, the results obtained can be extended to an arbitrary model of massless rqed@xmath7 with the help of an exact evaluation of @xmath143 , see app .
[ app : g(alpha , beta ) ] for general formulas .
we plan to return to this task , with application to the case of rqed@xmath198 which requires the knowledge of the complicated contribution @xmath143 , in our future investigations .
the work of a.v.k . was supported in part by the russian foundation for basic research ( grant no . 13 - 02 - 01005 ) and by the universit pierre et marie curie ( upmc ) .
here we give the expansion of the master integrals , or @xmath102-functions , contributing to the fermion self - energies in eqs .
( [ ren - sigmas])-([ren - sigma2s ] ) at fixed @xmath46 .
the results have the following form : [ exp : gfuncs ] & & g(1,1-_e)= ( 1+_1_+ ( _ 1 ^ 2 + 2_2 -3_2(1-_e ) ) ) , [ g1,1-ee ] + & & g(1,_-_e)= - ( 1+_1_+ ( _ 1 ^ 2 + 8_2 -9_2(1-_e ) ) ) , [ g1,eg - ee ] + & & g(1-_e,_)= - ( 1+_1_+ ( _ 1 ^ 2 + 4_2 -5_2(1-_e ) ) ) , [ g1-ee , eg ] + & & g(1,1)= ( 1+_2_+ ( _ 2 ^ 2+_2(_e)+2_2(1-_e)-4_2(2 - 2_e ) ) ) , where @xmath74 is the euler constant and @xmath175 , @xmath199 and @xmath200 were determined in ( [ ren - params - rqed4de ] ) .
in this appendix , we derive general expressions for the coefficient function @xmath201 , see eq .
( [ complicatedg ] ) . this function can be expressed in the following form : g(,1,,1,1 ) = g
( , , , , ) , a()= , = - , = -1 , [ complicatedg1 ] where g(,,,,)= c_d .
[ complicatedgt ] the function @xmath202 belongs to the type of feynman integrals considered in [ ] : it corresponds to the notation @xmath203 $ ] in [ ] and can be considered as the particular case where @xmath204 of the results in [ ] .
the general results contain four hypergeometric functions @xmath113 of argument @xmath53 .
we will show that the diagram @xmath202 contains only two hypergeometric functions @xmath113 of argument @xmath53 . following [ ]
, @xmath202 can be represented in the following form : ) , the factor @xmath205 , which appears explicitly in @xmath206 $ ] in [ ] , has already been extracted from the expression of @xmath102 . ]
g(,,,,)= ( , ) , ( , ) = ( , ) - ( , ) .
[ complicatedgt1 ] taking @xmath204 , the results ( 17 ) in [ ] becomes & & ( , ) = \{_n=0^ + & & + } , [ oi ] where we have used the property : @xmath207 $ ] for any @xmath24 .
note that the second term in the r.h.s . of ( 17 ) in [ ] has been summed in the product of @xmath174-functions , yielding the last term in eq .
( [ oi ] ) . for the part @xmath208
there are four representations , _
i.e. _ , the equations ( 18)-(22 ) in [ ]
. we use equation ( 21 ) , : in the denominator , @xmath209 should be replaced by @xmath210 . ] where the last terms is zero at @xmath211 .
so , we have & & ( , ) = + & & + _ n=0^ .
[ ti ] now , we consider the difference @xmath212 , _
i.e. _ , @xmath213 . the first term and the second one in the r.h.s . of ( [ oi ] ) and ( [ ti ] ) , respectively , combine to one term
. the terms without series also combine to one term .
so , we have & & ( , ) = -_n=0^ + & & - [ oi - ti ] it is convenient to transform the first term in the r.h.s .
to the new form containing the factor @xmath214 in its denominator .
it is possible to obtain from the transformation of @xmath113 hypergeometric functions of argument @xmath53 ( see eq . ( 9 ) in [ ] : _
n=0^ = - _ n=0^ , [ transform ] with arbitrary @xmath215 , @xmath216 , @xmath217 and @xmath218 . indeed , if @xmath219 , @xmath220 , @xmath221 and @xmath222 , we have & & _ n=0^ = + & & - _ n=0^ . [ transform1 ]
taking together this new results with the last two terms in eq .
( [ oi - ti ] ) , we have : ( , ) = i ( , ) , [ ht ] where & & i ( , ) = + _ n=0^ + & & ( - ) .
[ oi - tin ] the results ( [ ht ] ) and ( [ oi - tin ] ) together with ( [ complicatedg1 ] ) and ( [ complicatedgt ] ) can be considered as the final result for the initial diagram @xmath201 : g(,1,,1,1 ) = i ( , ) .
[ gn1 ] however , in the case where @xmath73 and @xmath89 are close to @xmath223 , it is not so convenient because ( [ oi - tin ] ) contains several additional singularities , which are canceled only at the end of calculations .
another form of the final result can be obtained by application of the transformation ( [ transform ] ) to the last term in eq .
( [ oi - tin ] ) . some simple algebra yields the following expression : & & i ( , ) = + _ n=0^ ( + & & + ) .
[ oi - tin1 ] it seems that the result ( [ oi - tin1 ] ) together with eq .
( [ gn1 ] ) is the most convenient final form of the result for the initial diagram @xmath201 .
there are two other forms of the result for @xmath224 , which can be obtained by application of the transform ( [ transform ] ) to the second term in eqs .
( [ oi - tin ] ) and ( [ oi - tin1 ] ) .
that leads to _
n=0^ = _ n=0^
_ 1(n+2 ) - _ 1(+-1 ) .
[ term ] using this formula the expressions of @xmath224 become more cumbersome because they contain the euler ( digamma ) @xmath162-functions . nevertheless , we also present them for completeness .
they are following & & i ( , ) = - _ 1(+-1 ) + & & + _ n=0^ _ 1(n+2 ) + & & - _ n=0^ .
[ oi - tin2 ] and & & i ( , ) = - _ 1(+-1 ) + & & + _ n=0^ ( + & & + _ 1(n+2 ) ) .
[ oi - tin3 ] unfortunately , the case @xmath225 ( _ i.e. _ , @xmath226 ) does not produce any additional simplification and , in this case , we should use the combination of the equations ( [ ht ] ) and ( [ oi - tin ] ) together with ( [ complicatedg1 ] ) and ( [ complicatedgt ] ) .
e. g. mishchenko , europhys .
* 83 * 17005 ( 2008 ) ; i. f. herbut , v. jurii and o. vafek , 046403 ( 2008 ) ; d. e. sheehy and j. schmalian , 193411 ( 2009 ) ; v. jurii , o. vafek and i. f. herbut , 235402 ( 2010 ) ; i. sodemann and m. m. fogler , 115408 ( 2012 ) .
n. a. kivel , a. s. stepenenko and a. n. vasilev , nucl .
* b 424 * 619 ( 1994 ) ; a. n. vasiliev , s. e. derkachov , n. a. kivel , and a. s. stepanenko , tmf * 94 * 179 ( 1993 ) [ theor . math .
phys . * 94 * 127 ( 1993 ) ] .
d. j. broadhurst , j. a. gracey , d. kreimer , z. phys .
* c 75 * 559 ( 1997 ) . | we compute the two - loop fermion self - energy in massless reduced quantum electrodynamics for an arbitrary gauge using the method of integration by parts . focusing on the limit where the photon field is four - dimensional
, our formula involves only recursively one - loop integrals and can therefore be evaluated exactly . from this formula
, we deduce the anomalous scaling dimension of the fermion field as well as the renormalized fermion propagator up to two loops .
the results are then applied to the ultra - relativistic limit of graphene and compared with similar results obtained for four - dimensional and three - dimensional quantum electrodynamics . |
the first stars are thought to form at @xmath5 , with the first galaxies following at @xmath6 ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the chemical abundances of these first galaxies are unknown .
if those abundances could be measured , then they would constrain the properties of metal - free population iii stars , the early chemical evolution of galaxies , and the reionization of the universe .
metal - poor stars in the milky way provide a local link to this high - redshift universe through the elemental abundances of their photospheres .
as the number of known metal - poor stars with detailed chemical abundance measurements has grown , it has become possible to homogeneously analyze large samples to search for subtle trends ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
it is tempting to assert that these metal - poor stars in the halo are the direct descendants of the first stars .
this is not necessarily the case though , as metal - poor stars form over a range of redshift in halos of varying mass and environment .
likewise , stars at a given redshift form with a range of metallicity .
the examination of other properties beyond metallicity are therefore necessary to identify the stars in the milky way that formed at the highest redshifts .
@xcite showed that because galaxies form from the inside - out , the oldest stars at a given metallicity are found near the center of a halo on the most tightly - bound orbits .
indeed , near the center of a milky way - analog a large fraction of stars with @xmath7 } \lesssim -2 $ ] formed at @xmath8 , while 2040% of stars with @xmath9 } \lesssim -3 $ ] formed at @xmath10 .
consequently , the metal - poor stellar population in the inner few kpc of the galaxy the bulge is the best place to search for truly ancient stars , including low - mass population iii stars that may have survived to the present day .
large - scale spectroscopic surveys of the bulge have shown that while metal - poor stars in the bulge are quite rare , they do exist .
the abundances and radial velocity galactic origins ( argos ) survey of @xcite and @xcite identified 16 stars with @xmath0 } \lesssim -2.0 $ ] in a sample of 14,150 stars within 3.5 kpc of the galactic center .
the most metal - poor star in their sample has @xmath0 } \approx -2.6 $ ] . as part of the third phase of the sloan digital sky survey , the apache point observatory galactic evolution experiment ( apogee ) collected @xmath11-band spectra for 2,403 giants stars in outer bulge fields and identified two stars with @xmath0 }
\approx -2.1 $ ] @xcite
. ground - based objective prism surveys for metal - poor stars in the bulge are impractical due to crowding and strong absolute and differential reddening .
for this reason , searches for metal - poor stars have historically avoided the inner regions of our own galaxy . recently
though , the extremely metal - poor bulge stars with aaomega ( embla ) survey has successfully used narrow - band skymapper @xmath12-band photometry @xcite in the h & k region to pre - select candidate metal - poor stars for follow - up spectroscopy . in a sample of more than 8,600 stars , @xcite found in excess of 300 stars with @xmath0 } \lesssim -2.0$]including four stars with @xmath13 } \lesssim -2.5 $ ] . still , strong absolute and significant differential reddening limits the efficiency of near uv based selections for metal - poor stars in the bulge and restricts their applicability to outer - bulge regions . in @xcite , we described a new technique to identify candidate metal - poor stars using only near - infrared 2mass and mid - infrared _ wise _ photometry @xcite .
our infrared selection is well suited to a search for metal - poor stars in the bulge , as it is minimally affected by crowding or reddening .
we found that more than 20% of the candidates selected with our infrared selection are genuine very metal - poor ( vmp ) stars with @xmath14 } \lesssim -2.0 $ ] .
another 2% of our candidates are genuine extremely metal - poor ( emp ) stars with @xmath15 } \lesssim -3.0 $ ] . in a sample of 90 metal - poor candidates selected with only an apparent magnitude cut to be high in the sky from las campanas in the first half of the year
we identified three stars with @xmath16 } \lesssim -2.7 $ ] within 4 kpc of the galactic center .
two of these stars are the most metal - poor stars in bulge in the literature , while the third is comparable to the most metal - poor star from @xcite . because these stars are both tightly bound to the galaxy and very metal - poor , they are likely to be among the most ancient stars identified to this point . for that reason ,
their detailed abundances provide clues to the chemistry of the first galaxies in the @xmath17 universe , beyond those already identified in more metal - poor halo stars .
these stars all have apparent magnitudes @xmath18 , making them unusually bright for stars at the distance of the bulge .
their bright apparent magnitudes enable a very telescope - time efficient exploration of the @xmath17 universe .
we describe the collection of the data we will subsequently analyze in section [ sec : data ] .
we detail the determination of distances and orbital properties , stellar parameters , and chemical abundances of these three stars in section [ sec : analysis ] .
we discuss our results and their implications in section [ sec : discussion ] , and we summarize our findings in section [ sec : conclusions ] .
we initially selected these stars as candidates according to criteria ( 1)(4 ) from section 2 of @xcite : @xmath19 , @xmath20 , @xmath21 , and @xmath22 .
we give astrometry and photometry for each star in table [ tbl-1 ] .
we confirmed their metal - poor nature using low - resolution spectroscopy from gemini south / gmos - s @xcite in service mode during march and april of 2014 .
our gemini south / gmos - s follow - up spectroscopy was not focused on candidates in the bulge , so the discovery of these stars in the bulge was not predetermined by our survey strategy .
we used the magellan inamori kyocera echelle ( mike ) spectrograph @xcite on the clay telescope at las campanas observatory on 2014 june 2122 to obtain high - resolution , high signal - to - noise ( s / n ) spectra suitable for a detailed chemical abundance analysis .
we observed all three stars in 05 seeing at airmass @xmath23 with exposure times in the range 390590 seconds .
the total exposure time for all three sources combined was less than 24 minutes . including overheads ,
our magellan / mike observations for all three stars were completed in about 30 minutes .
we used the 07 slit and the standard blue and red grating azimuths , yielding spectra between 332 nm and 915 nm with resolution @xmath24 in the blue and @xmath25 in the red .
the resultant spectra have s / n @xmath26 pixel@xmath27 at 400 nm and s / n @xmath28 pixel@xmath27 at 600 nm . to obtain proper motions for each star , we cross - matched with both the ucac4 and spm4 proper motion catalogs using ` topcat`@xcite .
we list both sets of proper motions for our sample in table [ tbl-2 ] .
we reduced the spectra using the ` carpy ` software package @xcite . we continuum - normalized individual echelle orders using spline functions before joining them to form a single contiguous spectrum .
we estimate line - of - sight radial velocities by cross - correlating each spectrum with a normalized rest - frame spectrum of the well - studied metal - poor giant star hd 122563 .
we use the measured radial velocities to place the spectra in the rest - frame of the star . to determine the distances between the sun and each star in our sample , we use the scaling relation @xmath29 taking their characteristic mass as @xmath30 ,
the bolometric luminosity @xmath31 of our stars can be approximated as @xmath32 we then use equation [ eq - lum ] and the stellar parameters from @xcite listed in table [ tbl-3 ] to determine @xmath31 .
we use a 10 gyr , @xmath33 } = -2.5}$ ] , and @xmath34 = + 0.4}$ ] dartmouth isochrone to convert @xmath31 into an absolute @xmath35-band magnitude @xmath36 @xcite . given the available photometry , @xmath35 is least affected by extinction .
we de - redden the observed @xmath35 magnitudes using the @xcite dust maps as updated in @xcite along with the @xcite infrared extinction law .
the distance modulus @xmath37 then yields @xmath38 , the approximate distance of each star from the sun .
assuming the distance to the galactic center is @xmath39 kpc ( e.g. , * ? ? ?
* ) , we can then compute @xmath40 , the approximate distance of each star from the galactic center .
we perform a monte carlo simulation to account for the random observational uncertainties in @xmath35 , @xmath41 , @xmath42 , @xmath43 , and @xmath44 .
we sample 10,000 realizations from the uncertainty distributions for each quantity and compute @xmath38 and @xmath40 for each realization .
we give both distance estimates and their random uncertainties in the first two columns of table [ tbl-4 ] .
all three stars have @xmath45 kpc .
we compute the galactic orbits of each star in our sample using the ` galpy ` code , with initial conditions set by the observed heliocentric radial velocities and proper motions in table [ tbl-2 ] and estimated @xmath38 values from table [ tbl-4 ] . following @xcite ,
we model the milky way s potential as the superposition of a miyamoto - nagai disk with a radial scale length of 4 kpc and a vertical scale height of 300 pc , a hernquist bulge with a scale radius of 600 pc , and a navarro - frenk - white halo with a scale length of 36 kpc @xcite .
we assume that the miyamoto - nagai disk , the hernquist bulge , and the navarro - frenk - white halo respectively contribute 60% , 5% , and 35% of the rotational support at the solar circle .
we integrate the orbits for 200 orbital periods and derive the pericenters @xmath46 , apocenters @xmath47 , and eccentricities @xmath48 .
we perform a monte carlo simulation to account for the random observational uncertainties in @xmath38 , @xmath49 , @xmath50 , and @xmath51 .
we sample 1,000 realizations from the uncertainty distributions for each quantity and use those data as input to an orbital integration . in an attempt to quantify the systematic uncertainties that result from the input proper motion measurements , we include in table [ tbl-4 ] orbital properties and uncertainties estimated using both ucac4 and spm4 proper motions .
we estimate stellar parameters by classical excitation and ionization balance using unblended and lines . following the process described in @xcite
, we measure equivalent widths of individual absorption lines from the rest - frame spectra by fitting gaussian profiles .
we visually inspect all lines for quality , and discard blended or low - significance measurements . for these analyses , we assume transitions are in local thermodynamic equilibrium ( lte ) and employ the plane - parallel 1d @xmath1-enhanced model atmospheres from @xcite .
we use the atomic data compiled by @xcite , the @xcite solar chemical composition , and the february 2013 version of moog to calculate line abundances and synthesize spectra @xcite .
we require four conditions to be simultaneously met for a converged set of stellar parameters : zero trend in line abundance with excitation potential , zero trend in line abundances with reduced equivalent width , equal mean and abundances , and that the mean [ /h ] abundance must match the input model atmosphere abundance [ m / h ] . in practice
we accepted solutions where the slopes had magnitudes less than @xmath52 and the absolute abundance differences were less than @xmath53 dex .
our estimated stellar parameters are provided in table [ tbl-3 ] . to verify our spectroscopically - derived effective temperatures ,
we calculate effective temperatures using color temperature relations for 2mass @xmath54 and apass/2mass @xmath55 colors .
we use the @xcite dust maps as updated by @xcite to account for reddening in both colors .
however , we find that our photometric temperatures are as much as @xmath56k hotter than our spectroscopically - derived quantities .
to explore the reason for this discrepancy , we also estimate effective temperatures by comparing the observed balmer lines with synthetic spectra from @xcite .
our analysis of the h-@xmath57 profile suggests effective temperatures between @xmath58k and 4800k for all three stars , in excellent agreement with our excitation - ionization balance measurements .
as we show qualitatively in figure [ fig01 ] , our observed spectra are very similar to the well - studied metal - poor giant star hd122563 . given our independent effective temperature estimates , and since hd122563 is a red giant branch star with @xmath59k , @xmath60 , and @xmath61 } = -2.64 $ ] @xcite , we are confident in our derived spectroscopic effective temperatures
. moreover , we observe repeated saturated interstellar d absorption lines in our data .
these lines are indicative of multiple optically - thick gas clouds along the line - of - sight , each with distinct velocities . for these reasons
, we assert that the discrepancy between photometric and spectroscopic temperatures is likely due to poorly - characterized reddening in the outer bulge region .
given the spectral resolution and s / n ratios of our data , we estimate that the uncertainties in our spectroscopically - derived stellar parameters are about @xmath62k in @xmath63 , 0.2dex in @xmath43 , 0.1dex in [ fe / h ] , and 0.1 km s@xmath27 in microturbulence ( @xmath64 ) . we note that our stellar parameters ( @xmath65 , @xmath43 , [ fe / h ] , @xmath64 ) would change if we used different model atmospheres or included a proper treatment of non - lte effects .
for metal - poor giants , the non - lte treatment would increase the mean line abundance by about 0.1dex and result in higher surface gravities for a given effective temperature .
as an example , @xcite reports a slightly cooler temperature and higher surface gravity for hd 122563 than we find for our three stars .
however , in that study @xmath65 and @xmath43 were not derived by excitation and ionization equilibrium .
instead , they were fixed by bolometric temperature and angular diameter measurements from @xcite . with the stellar parameters fixed , @xcite noted that hd 122563 showed the largest abundance imbalance of and lines in their sample .
this indicates the the application of the equilibrium method in lte tends towards a different set of stellar parameters . in figure [ fig02 ]
we plot our stars alongside giant star ( i.e. , @xmath66 ) comparison samples from @xcite and @xcite . although these authors estimated surface gravities directly from isochrones , our stellar parameters are comparable to their determinations .
consequently , we are confident of our stellar parameter estimates . our high - resolution , high s / n magellan / mike spectra allow us to measure the abundances of many light , odd - z , @xmath1 , fe - peak , and neutron - capture elements . for most elements , we determine individual line abundances from the measured equivalent widths of clean , unblended atomic lines .
we take a synthesis approach for molecular features ( e.g. , ch ) , doublets ( e.g. , li ) , or atomic transitions with significant hyperfine structure and/or isotopic splitting ( namely sc , v , mn , co , cu , ba , la , and eu ) .
we use molecular data ( ch ) from @xcite .
our hyperfine structure and isotopic splitting data come from @xcite for sc , v , mn , co , and cu , from @xcite for ba , and from @xcite for la and eu .
we assume standard solar system isotopic fractions as collated by @xcite .
we report our equivalent width measurements in table [ tbl-5 ] and our derived abundances in table [ tbl-6 ] .
we estimate lithium abundances through synthesis of the li doublet at @xmath676707 .
this feature is quite weak in our spectra .
however , the abundances we obtain are typical for stars at the tip of the red giant branch .
we synthesize the @xmath68-band molecular feature at @xmath674323 to estimate carbon abundances .
none of our stars are carbon enhanced by the @xcite definition of [ c / fe ] @xmath69 . on the other hand ,
one of our stars is carbon enhanced by the @xcite definition that takes stellar evolutionary effects into account . in either case , there is not much carbon present in the photospheres of our stars[c / fe ] ranges from @xmath70 in j183713 - 314109 to @xmath71 in j181503 - 375120 .
we measure potassium abundances from equivalent widths of the strong transitions at @xmath677664 and @xmath677698 . given the radial velocities of our targets , these lines were mostly separated from the telluric a - band feature near @xmath677600 .
we detected in all three stars and derive abundances from the strong @xmath675889 and @xmath675895 transitions .
we measure from the @xmath673961 feature .
all three stars appear @xmath1-enhanced ( mg , ti , si , and ca ) . on average ,
the @xmath1-element abundances of these three metal - poor stars in the bulge are similar to those observed in large samples of halo metal - poor giant stars ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
[ mg / fe ] varies between @xmath72 and @xmath73 , while [ ca / fe ] changes marginally from @xmath74 to @xmath75 .
however , in all stars we find that [ /fe ] and [ /fe ] are slightly lower than the other @xmath1-elements , between [ ti / fe]@xmath76 and @xmath77 ( figure [ fig03 ] ) . in all stars ,
the mean abundances of neutral and ionized ti transitions agree within 0.030.08dex .
we measure [ /fe ] abundances from the @xmath673905 transition , yielding [ /fe ] abundance ratios between @xmath78 and @xmath79 .
there are a large number of fe - peak transitions available in our spectra : , , & , , , , , and . while sc , v , cr , mn , co , ni , and zn are clearly measurable in all stars from multiple unblended lines , we do not detect in j155730 - 293922 or j183713 - 314109 .
instead , we provide upper limits for from the @xmath675105 transition .
we also report a low - significance detection of in j181503 - 375120 of [ /fe ] @xmath80 .
our fe - peak abundance ratios generally follow the mean halo abundance trends observed by other authors in giant stars of similar metallicity ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
we find that [ /fe ] , [ /fe ] , and [ /fe ] abundances are at the extremes of the abundance distribution observed in halo metal - poor giant stars .
we show this in figure [ fig05 ] and explore possible explanations for these observations in section [ sec : discussion ] .
we measure elemental abundances from the first ( sr and y ) and second ( ba ) neutron - capture peaks .
we do not detect eu or la in our targets , and therefore we report upper limits for these elements in table [ tbl-6 ] .
sr and y have a common nucleosynthetic pathway , and we observe comparable abundance ratios for these elements in all three stars . as we show in figure [ fig04 ] ,
all of our measured neutron - capture abundances are indistinguishable from the abundances observed in halo metal - poor giant stars @xcite .
the uncertainties in chemical abundances are dominated by systematics , principally due to the uncertainties in determining stellar parameters .
we vary the stellar parameters of each star by the estimated uncertainties and calculate the resulting change in abundances .
we give the sign and magnitude of these effects in table [ tbl-7 ] , along with the quadrature sum of systematic uncertainties . due to a lack of lines for some elements , we adopt a minimum random uncertainty of 0.1dex .
we estimate total uncertainties as the quadrature sum of random and systematic uncertainties , which we list in table [ tbl-7 ] . for uncertainties in [ x
/ fe ] abundance ratios ( e.g. , as shown in figures [ fig03][fig05 ] ) , we adopt the quadrature sum of the total uncertainties in [ x / h ] and [ /h ] .
our initial survey was not targeted at the bulge , so we are observing all three stars at random orbital phases .
since a star on a radial orbit spends most of its orbit near apocenter , there is a strong prior that we are observing all three stars close to apocenter . our estimated galactocentric distances and orbital parameters for the three stars listed in table [ tbl-4 ] securely place j155730 - 293922 and j183713 - 314109 in the bulge on tightly bound orbits . in both cases , the currently - observed galactocentric distances are consistent with the idea that both stars are near apocenter .
at the same time , the differences in proper motion reported by ucac4 and spm4 deviate by up to 3-@xmath81 .
it seems clear that the quoted random proper motion uncertainties are not representative of the total uncertainties including the contribution from systematics .
both stars have @xmath18 and have had their proper motions matched to the correct 2mass sources , so the discrepancy is not due to faintness or misidentification .
nevertheless , the range in proper motions reported by ucac4 and spm4 should be an approximation of the effect of the unreported systematic uncertainties . since both ucac4 and spm4 place j155730 - 293922 and j183713 - 314109 on tightly bound orbits
, there is no reason to reject the idea that they are indeed tightly bound .
we therefore argue that since j155730 - 293922 and j183713 - 314109 are metal - poor , located near the center of the galaxy , and on tightly bound orbits , they are likely to be truly ancient stars according to the analysis described in @xcite . on the other hand , the orbital parameters listed in table [ tbl-4 ] for the star j181503 - 375120 suggest that it may be a halo star on a very eccentric orbit .
both ucac4 and spm4 agree that @xmath82 mas yr@xmath27 with high significance , indicating a substantial transverse velocity at @xmath83 kpc .
the problem with that scenario is that j181503 - 375120 would spend only a tiny fraction of its orbit near where it is observed today , and we are therefore observing it at a special time .
there are two possible interpretations of this observation .
the first is that both ucac4 and spm4 have somehow overestimated the @xmath50 proper motion of j181503 - 375120 .
this can not easily be rejected .
though the ucac4 and spm4 proper motion measurements were produced independently , they both used the same blue spm plates for their first epoch astrometry . in that case
, the apparently large proper motion of j181503 - 375120 could be the result of an issue with the same blue spm plate .
moreover , both ucac4 and spm4 may be subject to residual systematic uncertainties at the level of 10 mas yr@xmath27 .
the second interpretation is that j181503 - 375120 is genuinely on a very eccentric orbit that takes it from the bulge all the way to the edge of the local group .
though we can not reject the latter hypothesis , we suspect that the former is a better explanation .
nevertheless , the proper motion of j181503 - 375120 merits further attention .
if its parallax is measured and its proper motion confirmed by gaia , then it could be a hypervelocity star that has been ejected from the galactic center by a three - body interaction involving the milky way s supermassive black hole .
in any case , j181503 - 375120 is currently located near the center of the galaxy .
since all three stars in our sample are old , one might wonder if the orbits we observe today might be significantly different from their orbits at higher redshift . even though we will argue that our stars formed at @xmath6 , they were likely accreted by the milky way more recently .
@xcite found that even metal - poor stars that formed at @xmath84 are not typically accreted by a milky way analog until @xmath85 .
@xcite showed that in the absence of a major merger , inside of 2 kpc milky way - analog dark matter halos have accreted more than 75% of their @xmath86 mass by @xmath85 .
the milky way is not likely to have had a major merger in that interval , as its disk is quite old and its bulge appears to be a psuedobulge best explained by secular disk instabilities ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the fact that the mass enclosed by the orbits of of our stars does not change much since they likely entered the milky way s dark matter halo suggests that their orbits should not have changed significantly
. the impact of merger activity would be to cause the outward diffusion of stellar orbits anyway , so in that situation the orbits of our stars would have been even more tightly bound in the past .
this would not qualitatively effect our interpretation of their abundances .
the inside - out formation of the milky way suggests that in the inner few kpc of the galaxy , about 10% of stars with @xmath0 } \lesssim -3.0 $ ] formed at @xmath5 @xcite .
another 2040% of stars in the range @xmath15 } \lesssim -3.0 $ ] formed at @xmath87 .
all three of our stars are currently in the inner galaxy , while the kinematics of two of the three place them on tightly bound orbits .
the probability @xmath88 that at least one of our stars formed at @xmath5 is 1 minus the probability that none of them formed at @xmath5 : @xmath89 .
likewise , the probability @xmath90 that at least one of our stars formed at @xmath17 is @xmath91 . in other words , there is 30% chance that at least one of these three stars formed at @xmath5 and a 70% chance that at least one star formed at @xmath92 .
if we apply the @xcite analysis only to j155730 - 293922 and j183713 - 314109 , then @xmath93 and @xmath94 . even though these stars are not the most metal - poor stars known , the combination of their low metallicity and tightly - bound orbits suggests that they may be among the most ancient stars with detailed chemical abundance measurements .
in this scenario , our derived chemical abundances are indicative of the chemical abundances of the progenitor galaxies of the milky way during the epoch of the first galaxies .
generally , we find that our abundance ratios are near the mean of abundance distributions observed in halo metal - poor giant stars @xcite .
si , sc and mn are exceptions though , which we discuss below .
based on four metal - poor bulge stars with @xmath95 } \lesssim -2.5 $ ] from the embla survey , @xcite reached a similar conclusion : bulge metal - poor stars have a similar abundance pattern to halo metal - poor stars .
they also noted large scatter in [ /fe ] from @xmath96 to @xmath97 in just four stars , with one star overabundant in [ /fe ] to the level of @xmath98 .
we find very little variance in [ /fe ] , ranging from @xmath72 to @xmath73 .
none of our stars are overabundant in either [ /fe ] , [ /fe ] , or any other @xmath1-elements .
in fact , we find that [ /fe ] and [ /fe ] are about @xmath99dex below the abundances of other @xmath1-elements .
our stars appear near the extremes of the silicon abundance distribution observed in halo metal - poor giant stars ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
this is likely due to their low surface gravities and temperatures though .
@xcite found that giants exhibited higher [ si / fe ] abundance ratios than dwarfs by about 0.2dex .
similarly , cool stars usually appear to have high silicon @xcite .
given these two effects , the slightly higher [ si / fe ] abundance ratios we find can most likely be attributed to a combination of low surface gravity and cool temperatures . indeed ,
when we consider [ si / fe ] in giant stars ( @xmath100 ) in the @xcite sample , our [ si / fe ] ratios lie near the mean for our temperature range .
that is to say although our stars show relatively high [ si / fe ] ratios , the stars with high [ si / fe ] values in the comparison samples also usually have cooler temperatures . in short , [ si / fe ]
appears to be strongly correlated with temperature . on the other hand ,
high silicon is consistent with the galactic chemical enrichment model predictions of @xcite . while we regard the former as the most likely explanation , we can not rule out the latter idea that the high silicon we observe is representative of the @xmath17 interstellar medium .
the [ /fe ] abundance ratios we find are lower than what is observed in metal - poor giants in the halo .
we use the same hyperfine structure data for as the referenced authors and derive abundances from common lines . @xcite and @xcite have noted that the resonance triplet at 403 nm yields systematically lower abundances than other neutral mn lines . for that reason ,
@xcite empirically corrected their triplet abundances by about + 0.3dex , which explains most of the discrepancy we observe .
@xcite made no corrections , and we still find our stars in the lower envelope of their [ /fe ] distribution .
the remaining difference in [ /fe ] is probably attributable to our stars being at the tip of the giant branch . in halo metal - poor giant stars , many authors have noted a positive trend in the @xmath101}$ ] plane .
in other words , lower [ mn / fe ] abundances are found in cooler giants @xcite .
all three stars have low scandium abundances , with [ /fe ] @xmath102 .
@xcite found a tight abundance relation between [ /h ] and [ /h ] in halo stars , which is suggestive of a common nucleosynthetic environment .
figure [ fig06 ] shows that our stars deviate significantly from this relation . unlike or ,
our low [ /fe ] abundance ratios can not be easily explained by correlations with @xmath42 .
@xcite found a slight slope ( @xmath103 ) in the relationship between @xmath42 and [ /fe ] , such that cooler stars have lower [ /fe ] abundance ratios .
the typical range of [ /fe ] they measure for cool stars is @xmath104 to @xmath105 though .
our measurements are substantially below this range , with [ /fe ] = @xmath106 to @xmath107 .
scandium probably remains the most discrepant element between galactic chemical evolution models and observations of metal - poor stars , as models typically under - predict sc abundances by a factor of ten .
for example , @xcite predict constant @xmath108 \sim -1}$ ] for metal - poor stars , roughly an order of magnitude lower than the observed values of [ sc / fe ] @xmath109 . the abundance ratios we find in the inner few kpc of the galaxy bring our stars far closer to these predictions .
however , advances in modeling are required for both abundance measurements ( e.g. , non - lte treatment , @xmath1103d@xmath111 photospheres ) and galactic chemical evolution models .
departures from local thermodynamic equilibrium or 3d effects will alter the inferred sc abundances , while increasing the @xmath1-rich freeze - out or delaying neutrino processes during explosive nucleosynthesis may be necessary to increase sc yields in chemical evolution models ( e.g. , * ? ? ?
we searched the saga database and the compilation of @xcite for other galactic giant stars with [ /fe ] @xmath102 .
that search returned three objects : bs 16929 - 005 , he 0533 - 5340 , and he 1207 - 3108 . while bs 16929 - 005 was reported by @xcite to have [ /fe ] @xmath112 , that measurement did not take into account the hyperfine structure that is known to be important for scandium abundance measurements ( e.g. , *
* ) . in comparison
, @xcite accounted for hyperfine structure in bs 16929 - 005 and found [ /fe ] @xmath113 .
we regard the latter measurement as more reliable .
@xcite found [ /fe ] @xmath114 for he 0533 - 5340 and @xcite found [ /fe ] @xmath115 for he 1207 - 3108 .
however , both he 0533 - 5340 and he
1207 - 3108 are among the rare class of iron - rich " metal - poor stars in which most [ x / fe ] abundances are sub - solar .
this is in contrast to typical metal - poor stars , which are usually enhanced in at least the @xmath1 elements .
the combination of low [ /fe ] and @xmath1 enhancement that we see in our three metal - poor giants in the bulge is unprecedented in any of the 381 metal - poor giant stars in the saga database with scandium abundance measurements .
these three stars are therefore unlike any other known star in the galaxy .
we also searched @xcite for metal - poor stars in dwarf galaxies with [ /fe ] @xmath102 .
we found two examples , one from @xcite in coma berenices ( sdss j122657 + 235611/comber - s3 ) and one from @xcite in carina ( car 3 ) .
car 3 is an iron - rich " metal - poor star , so we do not consider it further . that leaves sdss j122657 + 235611/comber - s3 with [ /fe ] @xmath116 as the only giant star known with a similar abundance pattern to our three metal - poor giants in the bulge .
coma berenices is an ultra - faint dwarf spheroidal ( dsph ) galaxy with a @xmath117-band absolute magnitude of only @xmath118 @xcite .
it is also one of the most ancient galaxies known .
indeed , @xcite found a mean age of @xmath119 gyr for coma berenices based on _ hubble space telescope _
advanced camera for surveys photometry of its resolved stellar population .
that made it the oldest galaxy in their sample .
the apparent chemical abundance similarity between the ancient dsph coma berenices and our three stars in the bulge supports both the conclusion that our three stars are among the most ancient stars in our galaxy and the idea that low [ /fe ] may be a chemical indicator of ancient stellar populations .
our detailed chemical abundance analysis has assumed that transition levels are in a state of lte .
it is well known that this assumption breaks down in the upper levels of stellar photospheres , where departures from lte can significantly alter the inferred elemental abundance .
the direction and magnitude of these abundance changes are dependent on stellar parameters , atomic number , ionization level , absorption depth ( i.e. , the strength of the transition ) , among other factors .
many authors have investigated the effects of abundance deviations due to lte departures in well - studied metal - poor giant stars that are comparable to our program stars , like hd122563 ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . for metal - poor giant stars like those analyzed here
, the abundance changes due to departures from lte will be the largest for , , and .
the change in is significantly negative .
a ` positive correction ' refers to a higher abundance after accounting for departures from lte . ] : @xmath120 @xmath121 , such that in figure [ fig03 ] we have shown uncorrected ( i.e. , lte ) abundances from @xcite for a fair comparison .
is expected to show the largest absolute change , with positive deviations up to about @xmath122dex .
similarly we can expect our abundances to increase by about @xmath123dex with the proper inclusion of lte departure coefficients . however , these corrections would be of the same approximate order and direction for the halo comparison samples .
therefore we assert that the abundance ratios we find in metal - poor stars in the bulge would persist in the lower tail of [ /fe ] abundance distribution observed in comparable halo stars .
all other species examined here have expected abundance deviations less than @xmath124dex , with the average magnitude being about @xmath125dex @xcite .
we note that systematic abundance differences can also be expected due to surface granulation and convection , complex features which can not be accounted for in our 1d models .
our observations indirectly suggest that the progenitor galaxies of the milky way had reached @xmath0 } \sim -3.0 $ ] with an abundance pattern comparable to metal - poor halo stars by @xmath6 .
the chemical state of high - redshift galaxies can be measured directly by observations of metal - poor damped ly@xmath1 systems ( dlas ) in absorption in the spectra of background quasars .
many authors have measured the column densities and relative abundances of h , c , n , o , al , si , and fe to @xmath126 . at higher redshift , c , o , mg , si , and fe have been measured in dlas at @xmath127 @xcite .
at @xmath128 , the abundances of one system has been bounded to be less than 1/1,000 solar @xcite .
where [ c / fe ] , [ o / fe ] , and [ si / fe ] have been measured in high - redshift dlas , it has been found that the average abundances are in good agreement with those observed in metal - poor stars : @xmath129 } \approx 0.15 \pm 0.03 $ ] , @xmath130 } \approx 0.40 \pm 0.01 $ ] , and @xmath131 } \approx 0.37 \pm 0.01 $ ] .
our stars in the bulge are likely ancient and are well matched by the observed abundances in dlas .
only 500 myr passes between @xmath6 and @xmath127 ( e.g. , * ? ? ?
* ) , so it seems plausible that the @xmath6 abundances as observed in our ancient stars ( after correcting for @xmath43 and @xmath42 effects ) are comparable to those directly observed at @xmath127 .
we have measured the detailed chemical abundances of the three metal - poor stars with @xmath0 } \lesssim -2.7 $ ] in the bulge that we discovered in @xcite .
two of these three stars are the most metal - poor stars in the bulge in the literature , while the third is comparable to the most metal - poor star identified in @xcite .
we have carefully estimated the galactocentric distances and orbits of all three stars . while we find that all three have @xmath45kpc , only j155730 - 293922 and j183713 - 314109 can be securely placed on tightly - bound orbits .
j181503 - 375120 may be a halo star on a very eccentric orbit that is only passing through the bulge . while ucac4 and spm4 proper motion measurements favor a very eccentric orbit , the orbit is so extreme that it may be more likely that there is an issue with the spm blue plate that provides the first epoch astrometry for both catalogs . when combined with their metal - poor nature , their proximity to the center of the galaxy and
their tightly - bound orbits indicate that these stars may be some of the most ancient objects yet identified .
we use the theoretical models of @xcite to estimate that there is a 30% chance that at least one of these stars formed at @xmath5 and a 70% chance that at least one formed at @xmath87 .
we therefore argue that the chemical abundances we observe in these metal - poor stars is representative of the chemical state of the interstellar medium in the progenitor galaxies of the milky way at @xmath6 .
compared to observations of metal - poor giant stars of similar effective temperatures found in the galactic halo , we find similar [ x / fe ] abundance ratios for most elements .
however , we observe [ /fe ] abundance ratios lower than reported in the halo by about @xmath132dex .
scandium remains the element with the largest discrepancy between what is observed in halo metal - poor stars and what is predicted from models of galactic chemical evolution .
interestingly , when compared to the values observed in halo metal - poor stars , our [ /fe ] abundances are closer to predictions for the chemical abundances of the first galaxies ( e.g. , * ? ? ?
for these reasons , the progenitor halos of the milky way likely reached @xmath0 } \sim -3.0 $ ] by @xmath6 .
their chemical abundances were probably very similar to those observed in halo metal - poor stars with the possible exception of sc , which we observe to be low in these ancient stars in the bulge .
we thank judith cohen , anna frebel , gerry gilmore , paul schechter , and josh winn .
we are especially grateful to the anonymous referee for suggestions that improved this paper .
this research has made use of nasa s astrophysics data system bibliographic services and both the simbad database and vizier catalog access tool , cds , strasbourg , france .
the original description of the vizier service was published by @xcite .
this research made use of astropy , a community - developed core python package for astronomy @xcite .
this publication makes use of data products from the wide - field infrared survey explorer , which is a joint project of the university of california , los angeles , and the jet propulsion laboratory / california institute of technology , funded by the national aeronautics and space administration .
this publication makes use of data products from the two micron all sky survey , which is a joint project of the university of massachusetts and the infrared processing and analysis center / california institute of technology , funded by the national aeronautics and space administration and the national science foundation .
this publication was partially based on observations obtained at the gemini observatory , which is operated by the association of universities for research in astronomy , inc .
, under a cooperative agreement with the nsf on behalf of the gemini partnership : the national science foundation ( united states ) , the national research council ( canada ) , conicyt ( chile ) , the australian research council ( australia ) , ministrio da cincia , tecnologia e inovao ( brazil ) and ministerio de ciencia , tecnologa e innovacin productiva ( argentina ) . 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 .
this research was made possible through the use of the aavso photometric all - sky survey ( apass ) , funded by the robert martin ayers sciences fund .
a. r. c. acknowledges support through european research council grant 320360 : the gaia - eso milky way survey .
support for this work was provided by the mit kavli institute for astrophysics and space research through a kavli postdoctoral fellowship .
abel , t. , bryan , g. l. , & norman , m. l.2002 , science , 295 , 93 anders , e. , & grevesse , n.1989 , , 53 , 197 andrievsky , s. m. , spite , m. , korotin , s. a. , et al .
2010 , , 509 , aa88 aoki , w. , beers , t. c. , christlieb , n. , et al .
2007 , , 655 , 492 asplund , m. , carlsson , m. , & botnen , a. v. 2003 , , 399 , l31 asplund , m. , grevesse , n. , sauval , a. j. , & scott , p. 2009
, , 47 , 481 astropy collaboration , robitaille , t. p. , & tollerud , e. j. , greenfield , p. , et al . 2013 , , 558 , 33 aumer , m. , & binney , j. j. 2009 , , 397 , 1286 barklem , p. s. , & piskunov , n.2003 , modelling of stellar atmospheres , 210 , 28p beers , t. c. , & christlieb , n.2005 , , 43 , 531 becker , g. d. , sargent , w. l. w. , rauch , m. , & carswell , r. f. 2012 , , 744 , 91 belokurov , v. , zucker , d. b. , evans , n. w. , et al .
2007 , , 654 , 897 bergemann , m. , & nordlander , t. 2014 , arxiv:1403.3088 bernstein , r. , shectman , s. a. , gunnels , s. m. , mochnacki , s. , & athey , a. e. 2003 , , 4841 , 1694 bessell , m. , bloxham , g. , schmidt , b. , et al .
2011 , , 123 , 789 bimont , e. , palmeri , p. , & quinet , p. 1999
, , 269 , 635 bonifacio , p. , spite , m. , cayrel , r. , et al .
2009 , , 501 , 519 bovy , j. 2015 , , 216 , 29 bovy , j. , hogg , d. w. , & rix , h .- w .
2009 , , 704 , 1704 bovy , j. , rix , h .- w . ,
liu , c. , et al .
2012 , , 753 , 148 bromm , v. , coppi , p. s. , & larson , r. b. 1999 , , 527 , l5 bromm , v. , & yoshida , n. 2011 , , 49 , 373 brown , t. m. , tumlinson , j. , geha , m. , et al .
2014 , , 796 , 91 casey , a. r. 2014 , arxiv:1405.5968 castelli , f. , & kurucz , r. l.2004 , arxiv : astro - ph/0405087 cayrel , r. , depagne , e. , spite , m. , et al .
2004 , , 416 , 1117 cohen , j. g. , christlieb , n. , thompson , i. , et al .
2013 , , 778 , 56 cooke , r. , pettini , m. , steidel , c. c. , rudie , g. c. , & jorgenson , r. a. 2011 , , 412 , 1047 cooke , r. , pettini , m. , steidel , c. c. , rudie , g. c. , & nissen , p. e. 2011 , , 417 , 1534 creevey , o. l. , thvenin , f. , boyajian , t. s. , et al .
2012 , , 545 , aa17 de jong , j. t. a. , rix , h .- w . ,
martin , n. f. , et al .
2008 , , 135 , 1361 demarque , p. , woo , j .- h . , kim , y .- c . , &
yi , s. k. 2004 , , 155 , 667 dessauges - zavadsky , m. , dodorico , s. , mcmahon , r. g. , et al .
2001 , , 370 , 426 dessauges - zavadsky , m. , proux , c. , kim , t .- s . , dodorico , s. , & mcmahon , r. g. 2003 , , 345 , 447 dotter , a. , chaboyer , b. , jevremovi , d. , et al .
2008 , , 178 , 89 ellison , s. l. , prochaska , j. x. , hennawi , j. , et al .
2010 , , 406 , 1435 fitzpatrick , e. l. 1999 , , 111 , 63 frebel , a. 2010 , astronomische nachrichten , 331 , 474 frebel , a. , simon , j. d. , geha , m. , & willman , b. 2010 , , 708 , 560 freeman , k. , ness , m. , wylie - de - boer , e. , et al .
2013 , , 428 , 3660 frhlich , c. , martnez - pinedo , g. , liebendrfer , m. , et al .
2006 , physical review letters , 96 , 142502 garca prez , a. e. , cunha , k. , shetrone , m. , et al . 2013 , , 767 , l9 girard , t. m. , van altena , w. f. , zacharias , n. , et al .
2011 , , 142 , 15 gratton , r. g. , carretta , e. , eriksson , k. , & gustafsson , b. 1999 , , 350 , 955 hansen , c. j. , bergemann , m. , cescutti , g. , et al .
2013 , , 551 , aa57 henden , a. a. , levine , s. e. , terrell , d. , smith , t. c. , & welch , d. 2012 , journal of the american association of variable star observers ( jaavso ) , 40 , 430 hernquist , l. 1990 , , 356 , 359 honda , s. , aoki , w. , ando , h. , et al .
2004 , , 152 , 113 hook , i. m. , jrgensen , i. , allington - smith , j. r. , et al .
2004 , , 116 , 425 howard , c. d. , rich , r. m. , clarkson , w. , et al .
2009 , , 702 , l153 howes , l. m. , asplund , m. , casey , a. r. , et al .
2014 , , 445 , 4241 indebetouw , r. , mathis , j. s. , babler , b. l. , et al .
2005 , , 619 , 931 jofr , p. , heiter , u. , soubiran , c. , et al .
2014 , , 564 , aa133 kelson , d. d. 2003 , , 115 , 688 kelson , d. d. , williams , r. j. , dressler , a. , et al .
2014 , , 783 , 110 kobayashi , c. , umeda , h. , nomoto , k. , tominaga , n. , & ohkubo , t. 2006 , , 653 , 1145 kormendy , j. , & kennicutt , r. c. , jr .
2004 , , 42 , 603 kurucz , r. , & bell , b. 1995 , atomic line data ( r.l .
kurucz and b. bell ) kurucz cd - rom no
cambridge , mass . : smithsonian astrophysical observatory , 1995 . , 23 lai , d. k. , bolte , m. , johnson , j. a. , et al .
2008 , , 681 , 1524 lawler , j. e. , & dakin , j. t. 1989 , journal of the optical society of america b optical physics , 6 , 1457 lawler , j. e. , bonvallet , g. , & sneden , c. 2001a , , 556 , 452 lawler , j. e. , wickliffe , m. e. , den hartog , e. a. , & sneden , c. 2001b , , 563 , 1075 mainzer , a. , bauer , j. , grav , t. , et al .
2011 , , 731 , 53 mashonkina , l. , zhao , g. , gehren , t. , et al .
2008 , , 478 , 529 masseron , t. , plez , b. , van eck , s. , et al .
2014 , , 571 , aa47 miyamoto , m. , & nagai , r.1975 , , 27 , 533 molaro , p. , bonifacio , p. , centurin , m. , et al .
2000 , , 541 , 54 navarro , j. f. , frenk , c. s. , & white , s. d. m. 1996 , , 462 , 563 ness , m. , freeman , k. , athanassoula , e. , et al .
2013 , , 432 , 2092 norris , j. e. , bessell , m. s. , yong , d. , et al . 2013a , , 762 , 25 norris , j. e. , yong , d. , bessell , m. s. , et al .
2013b , , 762 , 28 omeara , j. m. , burles , s. , prochaska , j. x. , et al .
2006 , , 649 , l61 ochsenbein , f. , bauer , p. , & marcout , j. 2000 , , 143 , 23 penprase , b. e. , prochaska , j. x. , sargent , w. l. w. , toro - martinez , i. , & beeler , d. j. 2010 , , 721 , 1 petitjean , p. , ledoux , c. , & srianand , r. 2008 , , 480 , 349 pettini , m. , zych , b. j. , steidel , c. c. , & chaffee , f. h. 2008 , , 385 , 2011 preston , g. w. , sneden , c. , thompson , i. b. , shectman , s. a. , & burley , g. s. 2006 , , 132 , 85 prochaska , j. x. , & mcwilliam , a. 2000 , , 537 , l57 prochaska , j. x. , & wolfe , a. m. 2002 , , 566 , 68 roederer , i. u. , preston , g. w. , thompson , i. b. , et al .
2014 , , 147 , 136 roederer , i. u. , sneden , c. , thompson , i. b. , preston , g. w. , & shectman , s. a. 2010 , , 711 , 573 schlafly , e. f. , & finkbeiner , d. p. 2011
, , 737 , 103 schlaufman , k. c. , & casey , a. r. 2014 , , 797 , 13 schlegel , d. j. , finkbeiner , d. p. , & davis , m. 1998 , , 500 , 525 schnrich , r. , & binney , j. 2009 , , 399 , 1145 shetrone , m. , venn , k. a. , tolstoy , e. , et al .
2003 , , 125 , 684 simcoe , r. a. , sullivan , p. w. , cooksey , k. l. , et al .
2012 , , 492 , 79 skrutskie , m. f. , cutri , r. m. , stiening , r. , et al .
2006 , , 131 , 1163 sneden , c. a. 1973 , phd thesis , univ . texas , austin sobeck , j. s. , kraft , r. p. , sneden , c. , et al .
2011 , , 141 , 175 srianand , r. , gupta , n. , petitjean , p. , noterdaeme , p. , & ledoux , c. 2010 , , 405 , 1888 suda , t. , katsuta , y. , yamada , s. , et al .
2008 , , 60 , 1159 suda , t. , yamada , s. , katsuta , y. , et al .
2011 , , 412 , 843 taylor , m. b. 2005 , in asp conf .
347 , astronomical data analysis software and systems xiv , ed .
p. shopbell , m. britton , & r. ebert ( san francisco , ca : asp ) , 29 tumlinson , j. 2010 , , 708 , 1398 wang , j. , navarro , j. f. , frenk , c. s. , et al .
2011 , , 413 , 1373 wright , e. l. 2006 , , 118 , 1711 wright , e. l. , eisenhardt , p. r. m. , mainzer , a. k. , et al .
2010 , , 140 , 1868 yamada , s. , suda , t. , komiya , y. , aoki , w. , & fujimoto , m. y. 2013 , , 436 , 1362 yong , d. , norris , j. e. , bessell , m. s. , et al . 2013a , , 762 , 26 yong , d. , norris , j. e. , bessell , m. s. , et al .
2013b , , 762 , 27 zacharias , n. , finch , c. t. , girard , t. m. , et al .
2013 , , 145 , 44 lccrrccccccc [email protected] & 15:57:30.1 & -29:39:23 & 344 & 18 & 13.15 & 0.97 & 11.13 & 10.61 & 10.50 & 10.39 & 10.40 + [email protected] & 18:15:03.6 & -37:51:20 & 355 & -10 & 12.87 & 1.01 & 10.80 & 10.29 & 10.19 & 10.12 & 10.13 + [email protected] & 18:37:13.2 & -31:41:09 & 3 & -11 & 12.47 & 0.82 & 10.65 & 10.12 & 10.04 & 9.90 & 9.90 lrrrrr [email protected] & @xmath134 & @xmath135 & @xmath136 & @xmath137 & @xmath138 + [email protected] & @xmath139 & @xmath140 & @xmath141 & @xmath142 & @xmath143 + [email protected] & @xmath144 & @xmath145 & @xmath146 & @xmath147 & @xmath148 lcccrccrc [email protected] & @xmath152 & @xmath153 & @xmath154 & @xmath155 & @xmath156 & @xmath157 & @xmath158 & @xmath159 + [email protected] & @xmath83 & @xmath160 & @xmath161 & @xmath162 & @xmath163 & @xmath164 & @xmath165 & @xmath166 + [email protected] & @xmath167 & @xmath168 & @xmath169 & @xmath170 & @xmath171 & @xmath172 & @xmath173 & @xmath174 cllrrrcrrcrr 5889.95 & & 0.00 & + 0.11 & 157.8 & 3.40 & & 170.4 & 3.52 & & 182.7 & 3.96 + 5895.92 & & 0.00 & @xmath1330.19 & 139.0 & 3.39 & & 144.0 & 3.41 & & 158.2 & 3.87 + 3829.36 & & 2.71 & @xmath1330.21 & 156.0 & 5.21 & & 169.6 & 5.32 & & & + 3832.30 & & 2.71 & + 0.27 & 191.9 & 5.13 & & & & & 197.3 & 5.30 + 3838.29 & & 2.72 & + 0.49 & 220.0 & 5.13 & & & & & 229.1 & 5.33 + 3986.75 & & 4.35 & @xmath1331.03 & 22.9 & 5.22 & & 40.8 & 5.59 & & 32.7 & 5.46 + 4057.51 & & 4.35 & @xmath1330.89 & & & & 34.5 & 5.31 & & 39.1 & 5.43 + 4167.27 & & 4.35 & @xmath1330.71 & 41.6 & 5.24 & & 55.8 & 5.47 & & 45.2 & 5.34 + 4571.10 & & 0.00 & @xmath1335.69 & 53.0 & 5.09 & & 72.4 & 5.36 & & 53.8 & 5.24 + 4702.99 & & 4.33 & @xmath1330.38 & 57.7 & 5.03 & & 73.6 & 5.25 & & 68.1 & 5.25 + 5172.68 & & 2.71 & @xmath1330.45 & 180.9 & 5.01 & & 198.3 & 5.17 & & 194.8 & 5.37 + 5183.60 & & 2.72 & @xmath1330.24 & 202.0 & 5.06 & & 216.6 & 5.17 & & 221.1 & 5.44 + 5528.40 & & 4.34 & @xmath1330.50 & 60.3 & 5.10 & & 77.7 & 5.33 & & 77.4 & 5.42 + 3961.52 & & 0.01 & @xmath1330.34 & 106.5 & 2.73 & & 126.6 & 3.07 & & 126.6 & 3.27 + 3905.52 & & 1.91 & @xmath1331.09 & 182.2 & 5.30 & & 211.0 & 5.53 & & 189.9 & 5.52 + 7664.90 & & 0.00 & + 0.14 & 43.0 & 2.54 & & 59.4 & 2.77 & & & + 7698.96 & & 0.00 & @xmath1330.17 & 38.1 & 2.77 & & 56.0 & 3.03 & & 43.5 & 2.94 + 4226.73 & & 0.00 & + 0.24 & 183.4 & 3.57 & & 212.1 & 3.82 & & 205.8 & 3.98 + 4283.01 & & 1.89 & @xmath1330.22 & 47.2 & 3.78 & & 59.9 & 3.98 & & 63.4 & 4.13 + 4318.65 & & 1.89 & @xmath1330.21 & 43.1 & 3.69 & & 57.2 & 3.91 & & 41.9 & 3.73 + 4425.44 & & 1.88 & @xmath1330.36 & 39.0 & 3.73 & & 52.6 & 3.95 & & 43.6 & 3.87 + 4435.69 & & 1.89 & @xmath1330.52 & 32.7 & 3.78 & & 43.1 & 3.96 & & 47.4 & 4.11 + 4454.78 & & 1.90 & + 0.26 & 68.5 & 3.62 & & 84.5 & 3.86 & & 82.8 & 3.97 + 4455.89 & & 1.90 & @xmath1330.53 & 30.4 & 3.75 & & 39.7 & 3.92 & & 39.6 & 3.99 + 5262.24 & & 2.52 & @xmath1330.47 & & & & 28.2 & 4.27 & & 28.8 & 4.34 + 5265.56 & & 2.52 & @xmath1330.26 & 25.2 & 4.00 & & 37.6 & 4.24 & & 38.9 & 4.32 + 5349.47 & & 2.71 & @xmath1330.31 & 11.4 & 3.85 & & 17.1 & 4.06 & & 17.9 & 4.13 + 5581.97 & & 2.52 & @xmath1330.56 & & & & & & & 21.5 & 4.24 + 5588.76 & & 2.52 & + 0.21 & 34.6 & 3.69 & & 45.7 & 3.87 & & 52.3 & 4.05 + 5590.12 & & 2.52 & @xmath1330.57 & & & & 18.2 & 4.11 & & 16.9 & 4.12 + 5594.47 & & 2.52 & + 0.10 & 27.0 & 3.65 & & 40.4 & 3.89 & & 41.9 & 3.99 + 5598.49 & & 2.52 & @xmath1330.09 & & & & 31.4 & 3.93 & & 34.7 & 4.05 + 5601.28 & & 2.53 & @xmath1330.52 & & & & 14.4 & 3.95 & & & + 5857.45 & & 2.93 & + 0.23 & 16.2 & 3.71 & & 23.7 & 3.92 & & 25.8 & 4.01 + 6102.72 & & 1.88 & @xmath1330.79 & 24.2 & 3.69 & & 38.3 & 3.95 & & 43.7 & 4.12 + 6122.22 & & 1.89 & @xmath1330.32 & 58.6 & 3.80 & & 67.2 & 3.91 & & 73.5 & 4.11 + 6162.17 & & 1.90 & @xmath1330.09 & 68.4 & 3.72 & & 75.6 & 3.81 & & 79.5 & 3.98 + 6169.06 & & 2.52 & @xmath1330.80 & 5.2 & 3.70 & & 12.7 & 4.13 & & 12.0 & 4.15 + 6169.56 & & 2.53 & @xmath1330.48 & 11.9 & 3.78 & & 17.0 & 3.96 & & 18.8 & 4.07 + 6439.07 & & 2.52 & + 0.47 & 45.1 & 3.55 & & 58.0 & 3.73 & & 59.5 & 3.84 + 6449.81 & & 2.52 & @xmath1330.50 & & & & 20.4 & 4.05 & & 17.2 & 4.02 + 4246.82 & & 0.32 & + 0.24 & & @xmath1330.70 & & & @xmath1330.50 & & & @xmath1330.15 + 4314.08 & & 0.62 & @xmath1330.10 & & @xmath1330.50 & & & @xmath1330.10 & & & 0.00 + 4325.00 & & 0.59 & @xmath1330.44 & & @xmath1330.40 & & & @xmath1330.10 & & & 0.00 + 4400.39 & & 0.61 & @xmath1330.54 & & @xmath1330.40 & & & @xmath1330.30 & & & @xmath1330.15 + 4415.54 & & 0.59 & @xmath1330.67 & & @xmath1330.25 & & & @xmath1330.25 & & & @xmath1330.15 + 5031.01 & & 1.36 & @xmath1330.40 & & & & & & & & @xmath1330.35 + 5526.78 & & 1.77 & + 0.02 & & @xmath1330.50 & & & @xmath1330.30 & & & @xmath1330.10 + 5657.91 & & 1.51 & @xmath1330.60 & & & & & @xmath1330.05 & & & 0.00 + 3904.78 & & 0.90 & + 0.03 & 20.7 & 2.45 & & & & & & + 3989.76 & & 0.02 & @xmath1330.06 & 57.3 & 2.16 & & 65.9 & 2.30 & & 67.7 & 2.48 + 3998.64 & & 0.05 & + 0.01 & 59.7 & 2.18 & & 63.9 & 2.23 & & 69.6 & 2.48 + 4008.93 & & 0.02 & @xmath1331.02 & 21.8 & 2.44 & & 20.5 & 2.41 & & 22.2 & 2.55 + 4512.73 & & 0.84 & @xmath1330.42 & & & & 11.6 & 2.39 & & 12.0 & 2.49 + 4518.02 & & 0.83 & @xmath1330.27 & 9.4 & 2.11 & & 21.1 & 2.53 & & & + 4533.25 & & 0.85 & + 0.53 & 38.6 & 2.09 & & 47.2 & 2.23 & & 49.2 & 2.38 + 4534.78 & & 0.84 & + 0.34 & 32.8 & 2.17 & & 38.3 & 2.27 & & 43.3 & 2.45 + 4535.57 & & 0.83 & + 0.12 & 22.0 & 2.15 & & 32.8 & 2.38 & & 32.4 & 2.47 + 4544.69 & & 0.82 & @xmath1330.52 & & & & & & & 12.9 & 2.60 + 4548.76 & & 0.83 & @xmath1330.30 & 12.5 & 2.28 & & 16.5 & 2.42 & & 17.2 & 2.53 + 4555.49 & & 0.85 & @xmath1330.43 & & & & & & & 12.1 & 2.51 + 4656.47 & & 0.00 & @xmath1331.29 & 11.7 & 2.22 & & 18.5 & 2.45 & & 16.6 & 2.50 + 4681.91 & & 0.05 & @xmath1331.01 & 20.9 & 2.28 & & 27.9 & 2.45 & & 23.2 & 2.46 + 4840.87 & & 0.90 & @xmath1330.45 & 11.6 & 2.43 & & 8.8 & 2.31 & & 11.2 & 2.51 + 4981.73 & & 0.84 & + 0.56 & 39.4 & 1.99 & & 60.1 & 2.30 & & 57.1 & 2.38 + 4991.07 & & 0.84 & + 0.44 & 44.1 & 2.18 & & 51.4 & 2.29 & & 53.4 & 2.44 + 4999.50 & & 0.83 & + 0.31 & 33.7 & 2.12 & & 46.8 & 2.34 & & 48.6 & 2.48 + 5007.21 & & 0.82 & + 0.17 & 38.6 & 2.33 & & 48.7 & 2.49 & & 59.6 & 2.78 + 5014.28 & & 0.81 & + 0.11 & & & & & & & 43.5 & 2.57 + 5016.16 & & 0.85 & @xmath1330.52 & 16.1 & 2.58 & & & & & & + 5020.02 & & 0.84 & @xmath1330.36 & 10.5 & 2.20 & & 19.3 & 2.51 & & 23.5 & 2.70 + 5024.84 & & 0.82 & @xmath1330.55 & 16.7 & 2.59 & & & & & & + 5036.46 & & 1.44 & + 0.19 & & & & 17.3 & 2.62 & & 7.5 & 2.29 + 5039.96 & & 0.02 & @xmath1331.13 & 17.8 & 2.23 & & 21.7 & 2.34 & & 24.6 & 2.52 + 5064.65 & & 0.05 & @xmath1330.94 & 26.2 & 2.28 & & 36.3 & 2.47 & & 40.4 & 2.66 + 5173.74 & & 0.00 & @xmath1331.06 & 19.7 & 2.17 & & 29.9 & 2.41 & & 20.7 & 2.32 + 5192.97 & & 0.02 & @xmath1330.95 & 28.3 & 2.28 & & 29.5 & 2.31 & & 24.6 & 2.32 + 5210.39 & & 0.05 & @xmath1330.83 & 25.5 & 2.14 & & 36.4 & 2.35 & & 30.3 & 2.36 + 3759.29 & & 0.61 & + 0.28 & 183.9 & 2.30 & & & & & & + 3761.32 & & 0.57 & + 0.18 & 185.0 & 2.36 & & & & & & + 3813.39 & & 0.61 & @xmath1332.02 & 89.4 & 2.74 & & 89.7 & 2.67 & & 93.1 & 2.84 + 3882.29 & & 1.12 & @xmath1331.71 & 44.6 & 2.16 & & 61.3 & 2.42 & & 61.9 & 2.47 + 3913.46 & & 1.12 & @xmath1330.42 & 109.5 & 2.13 & & 130.9 & 2.46 & & & + 4012.40 & & 0.57 & @xmath1331.75 & 96.3 & 2.40 & & 91.7 & 2.24 & & 111.4 & 2.76 + 4025.12 & & 0.61 & @xmath1331.98 & 70.7 & 2.18 & & 82.2 & 2.34 & & 84.8 & 2.48 + 4028.34 & & 1.89 & @xmath1330.96 & 41.9 & 2.24 & & 46.7 & 2.30 & & 55.6 & 2.46 + 4053.83 & & 1.89 & @xmath1331.21 & 31.1 & 2.27 & & 36.6 & 2.36 & & 41.4 & 2.46 + 4161.53 & & 1.08 & @xmath1332.16 & 34.7 & 2.28 & & 40.3 & 2.36 & & 51.6 & 2.58 + 4163.63 & & 2.59 & @xmath1330.40 & 32.2 & 2.29 & & 39.7 & 2.41 & & 47.9 & 2.56 + 4184.31 & & 1.08 & @xmath1332.51 & 22.0 & 2.36 & & 29.7 & 2.52 & & 33.0 & 2.60 + 4290.22 & & 1.16 & @xmath1330.93 & 93.2 & 2.09 & & 111.7 & 2.35 & & & + 4300.05 & & 1.18 & @xmath1330.49 & 106.6 & 1.92 & & 122.0 & 2.12 & & 124.2 & 2.33 + 4330.72 & & 1.18 & @xmath1332.06 & 30.6 & 2.18 & & 38.8 & 2.32 & & 45.3 & 2.45 + 4337.91 & & 1.08 & @xmath1330.96 & 98.6 & 2.09 & & 100.6 & 2.07 & & 112.6 & 2.41 + 4394.06 & & 1.22 & @xmath1331.78 & 48.0 & 2.23 & & 52.6 & 2.29 & & 60.7 & 2.45 + 4395.03 & & 1.08 & @xmath1330.54 & 115.4 & 1.96 & & 127.3 & 2.10 & & 131.5 & 2.34 + 4395.84 & & 1.24 & @xmath1331.93 & 34.3 & 2.18 & & 44.3 & 2.33 & & 50.9 & 2.46 + 4398.29 & & 1.21 & @xmath1332.65 & 8.6 & 2.14 & & & & & 15.0 & 2.41 + 4399.77 & & 1.24 & @xmath1331.19 & 76.5 & 2.11 & & 86.0 & 2.21 & & 93.2 & 2.44 + 4409.52 & & 1.23 & @xmath1332.37 & 19.6 & 2.29 & & 17.0 & 2.21 & & 22.3 & 2.36 + 4417.71 & & 1.17 & @xmath1331.19 & 80.0 & 2.07 & & 89.7 & 2.19 & & 93.6 & 2.35 + 4418.33 & & 1.24 & @xmath1331.97 & 31.8 & 2.17 & & 41.1 & 2.31 & & 45.0 & 2.40 + 4441.73 & & 1.18 & @xmath1332.41 & 21.6 & 2.31 & & 31.2 & 2.51 & & 37.4 & 2.64 + 4443.80 & & 1.08 & @xmath1330.72 & 104.4 & 1.91 & & 111.9 & 1.97 & & 122.1 & 2.32 + 4444.55 & & 1.12 & @xmath1332.24 & 27.6 & 2.20 & & 36.2 & 2.35 & & 48.3 & 2.57 + 4450.48 & & 1.08 & @xmath1331.52 & 72.1 & 2.15 & & 80.9 & 2.25 & & 90.7 & 2.51 + 4464.45 & & 1.16 & @xmath1331.81 & 50.5 & 2.21 & & 55.5 & 2.27 & & 64.2 & 2.43 + 4468.52 & & 1.13 & @xmath1330.60 & 108.9 & 1.92 & & 117.7 & 2.00 & & 128.4 & 2.36 + 4470.85 & & 1.17 & @xmath1332.02 & 33.5 & 2.15 & & 39.6 & 2.24 & & 49.3 & 2.43 + 4488.34 & & 3.12 & @xmath1330.82 & & & & & & & 11.3 & 2.68 + 4493.52 & & 1.08 & @xmath1333.02 & 12.1 & 2.50 & & 20.0 & 2.75 & & 18.7 & 2.72 + 4501.27 & & 1.12 & @xmath1330.77 & 105.1 & 2.00 & & 106.3 & 1.94 & & 117.3 & 2.29 + 4529.48 & & 1.57 & @xmath1332.03 & 32.2 & 2.61 & & 32.5 & 2.61 & & 43.4 & 2.81 + 4533.96 & & 1.24 & @xmath1330.53 & 109.3 & 1.97 & & 109.5 & 1.91 & & 125.9 & 2.35 + 4545.14 & & 1.13 & @xmath1331.81 & 18.9 & 1.56 & & & & & 26.6 & 1.75 + 4563.77 & & 1.22 & @xmath1330.96 & 97.6 & 2.16 & & 101.9 & 2.17 & & 113.7 & 2.50 + 4571.97 & & 1.57 & @xmath1330.32 & 92.8 & 1.86 & & 100.8 & 1.94 & & 114.6 & 2.30 + 4583.41 & & 1.16 & @xmath1332.92 & & & & 13.3 & 2.52 & & 16.2 & 2.63 + 4589.91 & & 1.24 & @xmath1331.79 & 54.5 & 2.32 & & 63.8 & 2.43 & & 70.7 & 2.58 + 4636.32 & & 1.16 & @xmath1333.02 & & & & 9.6 & 2.46 & & & + 4657.20 & & 1.24 & @xmath1332.24 & 20.7 & 2.15 & & 29.8 & 2.34 & & 33.0 & 2.42 + 4708.66 & & 1.24 & @xmath1332.34 & 22.2 & 2.28 & & 26.6 & 2.37 & & 32.7 & 2.51 + 4779.98 & & 2.05 & @xmath1331.37 & 20.2 & 2.23 & & 28.9 & 2.42 & & 34.1 & 2.52 + 4798.53 & & 1.08 & @xmath1332.68 & 18.5 & 2.32 & & 22.5 & 2.41 & & 30.6 & 2.60 + 4805.09 & & 2.06 & @xmath1331.10 & 35.8 & 2.29 & & 41.5 & 2.37 & & 51.7 & 2.55 + 4865.61 & & 1.12 & @xmath1332.81 & 15.9 & 2.41 & & 19.6 & 2.51 & & 17.9 & 2.48 + 4911.18 & & 3.12 & @xmath1330.34 & & & & & & & 13.2 & 2.22 + 5129.16 & & 1.89 & @xmath1331.24 & 27.4 & 2.03 & & 43.9 & 2.30 & & 47.5 & 2.37 + 5185.90 & & 1.89 & @xmath1331.49 & 26.8 & 2.26 & & 34.0 & 2.38 & & 35.3 & 2.42 + 5188.69 & & 1.58 & @xmath1331.05 & 72.7 & 2.14 & & 83.3 & 2.26 & & 83.5 & 2.33 + 5226.54 & & 1.57 & @xmath1331.26 & 55.6 & 2.10 & & 66.0 & 2.22 & & 67.2 & 2.29 + 5336.79 & & 1.58 & @xmath1331.59 & 41.7 & 2.22 & & 46.7 & 2.29 & & 52.1 & 2.39 + 5381.02 & & 1.57 & @xmath1331.92 & 20.3 & 2.13 & & 35.5 & 2.43 & & 38.0 & 2.49 + 5418.77 & & 1.58 & @xmath1332.00 & 19.6 & 2.20 & & 18.8 & 2.17 & & 29.4 & 2.42 + 4379.24 & & 2.10 & @xmath1332.71 & & 0.46 & & & 0.58 & & & 0.68 + 3916.41 & & 1.43 & @xmath1331.05 & 19.0 & 1.21 & & 24.5 & 1.34 & & 25.3 & 1.36 + 3951.96 & & 1.48 & @xmath1330.78 & 28.4 & 1.22 & & 38.3 & 1.39 & & 31.8 & 1.28 + 4005.71 & & 1.82 & @xmath1330.52 & & & & 32.5 & 1.42 & & 35.0 & 1.47 + 4023.38 & & 1.80 & @xmath1330.69 & 17.2 & 1.21 & & 22.4 & 1.34 & & 27.7 & 1.46 + 4035.62 & & 1.79 & @xmath1330.77 & & & & 41.5 & 1.78 & & & + 3908.76 & & 1.00 & @xmath1331.00 & & & & & & & 19.6 & 2.83 + 4254.33 & & 0.00 & @xmath1330.11 & 98.1 & 2.04 & & 99.1 & 1.99 & & 107.7 & 2.41 + 4274.80 & & 0.00 & @xmath1330.22 & 93.6 & 2.05 & & 101.2 & 2.13 & & 109.9 & 2.56 + 4289.72 & & 0.00 & @xmath1330.37 & 85.6 & 2.03 & & 99.7 & 2.25 & & 107.8 & 2.66 + 4337.57 & & 0.97 & @xmath1331.11 & & & & 24.9 & 2.85 & & & + 4545.95 & & 0.94 & @xmath1331.37 & & & & & & & 11.4 & 2.72 + 4580.06 & & 0.94 & @xmath1331.65 & & & & 15.7 & 3.06 & & & + 4600.75 & & 1.00 & @xmath1331.26 & & & & 11.1 & 2.57 & & & + 4616.14 & & 0.98 & @xmath1331.19 & 15.9 & 2.64 & & 13.5 & 2.57 & & 14.5 & 2.69 + 4626.19 & & 0.97 & @xmath1331.32 & & & & 17.2 & 2.81 & & & + 4646.15 & & 1.03 & @xmath1330.74 & 23.0 & 2.44 & & 29.2 & 2.58 & & 33.4 & 2.75 + 4651.28 & & 0.98 & @xmath1331.46 & & & & 8.7 & 2.62 & & 6.3 & 2.56 + 4652.16 & & 1.00 & @xmath1331.03 & 16.0 & 2.50 & & 18.3 & 2.58 & & 23.3 & 2.80 + 5206.04 & & 0.94 & + 0.02 & 67.9 & 2.24 & & 80.4 & 2.40 & & 85.1 & 2.65 + 5208.42 & & 0.94 & + 0.16 & 90.4 & 2.45 & & 96.9 & 2.52 & & & + 5247.56 & & 0.96 & @xmath1331.64 & & & & & & & 8.4 & 2.78 + 5296.69 & & 0.98 & @xmath1331.36 & 8.6 & 2.43 & & 11.0 & 2.56 & & 12.6 & 2.71 + 5298.28 & & 0.98 & @xmath1331.14 & & & & 22.5 & 2.69 & & 24.2 & 2.82 + 5345.80 & & 1.00 & @xmath1330.95 & 16.3 & 2.35 & & 31.4 & 2.70 & & 24.7 & 2.67 + 5348.31 & & 1.00 & @xmath1331.21 & 22.5 & 2.77 & & 12.0 & 2.46 & & 12.6 & 2.58 + 5409.77 & & 1.03 & @xmath1330.67 & 25.0 & 2.32 & & 34.3 & 2.50 & & 40.0 & 2.70 + 4558.59 & & 4.07 & @xmath1330.66 & 10.7 & 2.74 & & 16.3 & 2.94 & & 24.4 & 3.12 + 4588.14 & & 4.07 & @xmath1330.83 & 9.2 & 2.83 & & 14.0 & 3.02 & & 17.2 & 3.09 + 4030.75 & & 0.00 & @xmath1330.48 & & 1.51 & & & 1.68 & & & 1.93 + 4033.06 & & 0.00 & @xmath1330.62 & & 1.36 & & & 1.63 & & & 1.88 + 4034.48 & & 0.00 & @xmath1330.81 & & 1.46 & & & 1.63 & & & 1.78 + 4041.36 & & 2.11 & + 0.28 & & 1.66 & & & 1.88 & & & 2.18 + 3689.46 & & 2.94 & @xmath1330.17 & & & & 57.6 & 4.60 & & 54.1 & 4.62 + 3753.61 & & 2.18 & @xmath1330.89 & 58.9 & 4.43 & & 60.9 & 4.42 & & 65.6 & 4.65 + 3765.54 & & 3.24 & + 0.48 & 65.7 & 4.49 & & 70.7 & 4.54 & & 75.7 & 4.79 + 3786.68 & & 1.01 & @xmath1332.19 & & & & 82.9 & 4.74 & & 82.5 & 4.91 + 3805.34 & & 3.30 & + 0.31 & 58.2 & 4.53 & & 61.4 & 4.56 & & 63.4 & 4.70 + 3839.26 & & 3.05 & @xmath1330.33 & 50.8 & 4.68 & & 55.3 & 4.74 & & 42.9 & 4.58 + 3845.17 & & 2.42 & @xmath1331.39 & 28.0 & 4.51 & & & & & 33.8 & 4.71 + 3846.80 & & 3.25 & @xmath1330.02 & 40.9 & 4.40 & & 60.5 & 4.77 & & 46.4 & 4.57 + 3850.82 & & 0.99 & @xmath1331.75 & & & & 104.6 & 4.71 & & & + 3852.57 & & 2.18 & @xmath1331.18 & 56.3 & 4.59 & & 63.9 & 4.70 & & 65.8 & 4.88 + 3863.74 & & 2.69 & @xmath1331.43 & 19.6 & 4.66 & & 28.2 & 4.87 & & 34.5 & 5.08 + 3867.22 & & 3.02 & @xmath1330.45 & 27.8 & 4.28 & & 43.8 & 4.60 & & 55.5 & 4.91 + 3885.51 & & 2.42 & @xmath1331.09 & 41.6 & 4.48 & & 38.9 & 4.42 & & 42.1 & 4.56 + 3887.05 & & 0.91 & @xmath1331.14 & 110.7 & 4.20 & & & & & & + 3902.95 & & 1.56 & @xmath1330.44 & 117.4 & 4.45 & & 119.5 & 4.38 & & & + 3917.18 & & 0.99 & @xmath1332.15 & 79.3 & 4.55 & & 93.6 & 4.79 & & 96.5 & 5.07 + 3940.88 & & 0.96 & @xmath1332.60 & 49.8 & 4.35 & & 69.8 & 4.69 & & 67.6 & 4.81 + 3949.95 & & 2.18 & @xmath1331.25 & 59.3 & 4.66 & & 69.9 & 4.83 & & 61.6 & 4.79 + 3977.74 & & 2.20 & @xmath1331.12 & 66.1 & 4.67 & & 67.4 & 4.66 & & 63.8 & 4.72 + 4001.66 & & 2.18 & @xmath1331.90 & 21.5 & 4.53 & & 31.9 & 4.76 & & 27.2 & 4.74 + 4005.24 & & 1.56 & @xmath1330.58 & 112.8 & 4.37 & & 118.9 & 4.40 & & & + 4007.27 & & 2.76 & @xmath1331.28 & 19.5 & 4.54 & & 26.5 & 4.71 & & 27.3 & 4.80 + 4014.53 & & 3.05 & @xmath1330.59 & & & & & & & 64.4 & 5.20 + 4021.87 & & 2.76 & @xmath1330.73 & 39.7 & 4.43 & & 47.1 & 4.56 & & 55.2 & 4.80 + 4032.63 & & 1.49 & @xmath1332.38 & 26.9 & 4.30 & & 48.8 & 4.70 & & 49.4 & 4.83 + 4044.61 & & 2.83 & @xmath1331.22 & 18.2 & 4.52 & & 26.3 & 4.72 & & 28.3 & 4.84 + 4058.22 & & 3.21 & @xmath1331.11 & & & & 19.4 & 4.89 & & 16.6 & 4.86 + 4062.44 & & 2.85 & @xmath1330.86 & 34.3 & 4.55 & & 37.3 & 4.60 & & 42.1 & 4.77 + 4067.27 & & 2.56 & @xmath1331.42 & 24.2 & 4.55 & & 38.9 & 4.84 & & 36.0 & 4.87 + 4067.98 & & 3.21 & @xmath1330.47 & 28.5 & 4.47 & & 38.1 & 4.65 & & 36.5 & 4.69 + 4070.77 & & 3.24 & @xmath1330.79 & 14.8 & 4.45 & & 18.6 & 4.58 & & 23.8 & 4.77 + 4073.76 & & 3.27 & @xmath1330.90 & & & & 21.2 & 4.79 & & 26.7 & 4.98 + 4076.63 & & 3.21 & @xmath1330.37 & 41.4 & 4.62 & & 52.9 & 4.81 & & 52.5 & 4.90 + 4079.84 & & 2.86 & @xmath1331.36 & 10.7 & 4.41 & & 19.3 & 4.72 & & 21.0 & 4.83 + 4095.97 & & 2.59 & @xmath1331.48 & 16.0 & 4.41 & & 25.2 & 4.66 & & 29.9 & 4.84 + 4098.18 & & 3.24 & @xmath1330.88 & 18.8 & 4.66 & & 23.5 & 4.79 & & 23.4 & 4.84 + 4109.80 & & 2.85 & @xmath1330.94 & 29.7 & 4.52 & & 42.1 & 4.76 & & 34.0 & 4.68 + 4114.44 & & 2.83 & @xmath1331.30 & 13.0 & 4.40 & & 24.0 & 4.73 & & 22.3 & 4.76 + 4120.21 & & 2.99 & @xmath1331.27 & 19.8 & 4.78 & & 22.1 & 4.84 & & 27.2 & 5.03 + 4121.80 & & 2.83 & @xmath1331.45 & & & & 28.6 & 4.98 & & 24.3 & 4.95 + 4132.06 & & 1.61 & @xmath1330.68 & 115.3 & 4.48 & & 119.8 & 4.47 & & & + 4132.90 & & 2.85 & @xmath1331.01 & 28.8 & 4.56 & & 41.4 & 4.80 & & 40.5 & 4.87 + 4134.68 & & 2.83 & @xmath1330.65 & 39.9 & 4.40 & & 56.2 & 4.68 & & 54.4 & 4.74 + 4137.00 & & 3.42 & @xmath1330.45 & 17.9 & 4.41 & & 22.2 & 4.52 & & 26.1 & 4.68 + 4139.93 & & 0.99 & @xmath1333.63 & 12.1 & 4.49 & & 25.8 & 4.89 & & 18.5 & 4.81 + 4143.41 & & 3.05 & @xmath1330.20 & 52.6 & 4.44 & & & & & 65.5 & 4.77 + 4143.87 & & 1.56 & @xmath1330.51 & 117.4 & 4.28 & & 120.4 & 4.25 & & 130.5 & 4.72 + 4147.67 & & 1.48 & @xmath1332.07 & 55.9 & 4.47 & & 65.6 & 4.62 & & 70.7 & 4.86 + 4152.17 & & 0.96 & @xmath1333.23 & 34.6 & 4.62 & & 48.1 & 4.86 & & 45.5 & 4.94 + 4153.90 & & 3.40 & @xmath1330.32 & 29.4 & 4.53 & & 37.4 & 4.69 & & 41.5 & 4.84 + 4154.50 & & 2.83 & @xmath1330.69 & 43.5 & 4.50 & & 50.8 & 4.62 & & 49.1 & 4.68 + 4154.81 & & 3.37 & @xmath1330.40 & 25.9 & 4.50 & & 38.0 & 4.74 & & 31.7 & 4.69 + 4156.80 & & 2.83 & @xmath1330.81 & 43.6 & 4.62 & & 45.4 & 4.64 & & 47.7 & 4.77 + 4157.78 & & 3.42 & @xmath1330.40 & 20.2 & 4.42 & & 33.2 & 4.71 & & 35.7 & 4.83 + 4158.79 & & 3.43 & @xmath1330.67 & 12.4 & 4.44 & & 16.1 & 4.58 & & 18.7 & 4.71 + 4174.91 & & 0.91 & @xmath1332.94 & 55.1 & 4.63 & & 68.5 & 4.83 & & 58.9 & 4.81 + 4175.64 & & 2.85 & @xmath1330.83 & 32.1 & 4.44 & & 48.8 & 4.74 & & 43.9 & 4.74 + 4181.76 & & 2.83 & @xmath1330.37 & 54.6 & 4.37 & & 65.4 & 4.53 & & 70.9 & 4.77 + 4182.38 & & 3.02 & @xmath1331.18 & 12.1 & 4.45 & & 16.0 & 4.60 & & 20.2 & 4.79 + 4184.89 & & 2.83 & @xmath1330.87 & 34.0 & 4.49 & & 39.7 & 4.60 & & 36.7 & 4.62 + 4187.04 & & 2.45 & @xmath1330.51 & 75.2 & 4.43 & & 82.7 & 4.52 & & 83.6 & 4.71 + 4187.80 & & 2.42 & @xmath1330.51 & 74.4 & 4.37 & & 87.1 & 4.57 & & 86.7 & 4.74 + 4191.43 & & 2.47 & @xmath1330.67 & 65.9 & 4.43 & & 72.9 & 4.53 & & 77.8 & 4.76 + 4195.33 & & 3.33 & @xmath1330.49 & 27.9 & 4.58 & & 41.0 & 4.83 & & 33.5 & 4.76 + 4196.21 & & 3.40 & @xmath1330.70 & 18.4 & 4.63 & & & & & 18.8 & 4.71 + 4199.10 & & 3.05 & + 0.16 & 62.3 & 4.23 & & 74.4 & 4.43 & & 74.9 & 4.57 + 4202.03 & & 1.49 & @xmath1330.69 & 113.6 & 4.25 & & 121.7 & 4.33 & & 124.3 & 4.64 + 4216.18 & & 0.00 & @xmath1333.36 & 96.9 & 4.69 & & 99.8 & 4.68 & & 97.5 & 4.87 + 4217.55 & & 3.43 & @xmath1330.48 & 24.7 & 4.61 & & 30.6 & 4.73 & & 28.9 & 4.76 + 4222.21 & & 2.45 & @xmath1330.91 & 47.3 & 4.31 & & 61.6 & 4.53 & & 64.9 & 4.72 + 4227.43 & & 3.33 & + 0.27 & 66.7 & 4.54 & & 82.8 & 4.80 & & 91.2 & 5.13 + 4233.60 & & 2.48 & @xmath1330.58 & 67.2 & 4.36 & & 74.6 & 4.46 & & 78.8 & 4.69 + 4238.81 & & 3.40 & @xmath1330.23 & 30.7 & 4.45 & & 41.7 & 4.65 & & 41.9 & 4.73 + 4247.43 & & 3.37 & @xmath1330.24 & 39.5 & 4.59 & & 56.0 & 4.87 & & 56.5 & 4.97 + 4250.12 & & 2.47 & @xmath1330.38 & 77.0 & 4.32 & & 86.9 & 4.47 & & 88.4 & 4.68 + 4250.79 & & 1.56 & @xmath1330.71 & 105.9 & 4.15 & & 116.9 & 4.29 & & 121.0 & 4.64 + 4260.47 & & 2.40 & + 0.08 & 105.8 & 4.37 & & 108.1 & 4.33 & & 120.0 & 4.81 + 4271.15 & & 2.45 & @xmath1330.34 & & & & 103.0 & 4.71 & & 99.1 & 4.83 + 4282.40 & & 2.18 & @xmath1330.78 & 72.3 & 4.27 & & 81.0 & 4.40 & & 88.0 & 4.71 + 4337.05 & & 1.56 & @xmath1331.70 & 80.0 & 4.56 & & 85.1 & 4.60 & & 86.2 & 4.82 + 4352.73 & & 2.22 & @xmath1331.29 & 58.7 & 4.57 & & 70.7 & 4.74 & & 62.8 & 4.74 + 4375.93 & & 0.00 & @xmath1333.00 & 110.3 & 4.51 & & & & & 116.3 & 4.83 + 4388.41 & & 3.60 & @xmath1330.68 & & & & 12.3 & 4.61 & & 19.6 & 4.90 + 4404.75 & & 1.56 & @xmath1330.15 & 136.4 & 4.13 & & & & & & + 4407.71 & & 2.18 & @xmath1331.97 & 34.7 & 4.77 & & 45.7 & 4.96 & & 45.4 & 5.06 + 4415.12 & & 1.61 & @xmath1330.62 & 112.2 & 4.15 & & 121.3 & 4.25 & & 123.2 & 4.55 + 4422.57 & & 2.85 & @xmath1331.11 & 31.0 & 4.64 & & 34.4 & 4.70 & & 41.8 & 4.92 + 4427.31 & & 0.05 & @xmath1332.92 & 102.2 & 4.30 & & 118.9 & 4.57 & & 115.4 & 4.76 + 4430.61 & & 2.22 & @xmath1331.66 & 28.7 & 4.39 & & 46.1 & 4.70 & & 45.7 & 4.80 + 4442.34 & & 2.20 & @xmath1331.23 & 54.1 & 4.38 & & 67.0 & 4.57 & & 70.3 & 4.76 + 4443.19 & & 2.86 & @xmath1331.04 & 21.4 & 4.36 & & 33.1 & 4.62 & & 35.3 & 4.74 + 4447.72 & & 2.22 & @xmath1331.34 & 45.2 & 4.37 & & 65.4 & 4.67 & & 57.1 & 4.67 + 4454.38 & & 2.83 & @xmath1331.30 & 15.4 & 4.42 & & 19.9 & 4.55 & & & + 4459.12 & & 2.18 & @xmath1331.28 & 63.3 & 4.55 & & 82.2 & 4.85 & & 81.3 & 4.99 + 4461.65 & & 0.09 & @xmath1333.19 & 95.0 & 4.47 & & 105.6 & 4.60 & & 100.9 & 4.75 + 4466.55 & & 2.83 & @xmath1330.60 & 60.6 & 4.61 & & 69.2 & 4.73 & & 70.9 & 4.88 + 4476.02 & & 2.85 & @xmath1330.82 & 49.6 & 4.67 & & 55.6 & 4.75 & & 62.1 & 4.97 + 4484.22 & & 3.60 & @xmath1330.86 & 10.0 & 4.67 & & 10.0 & 4.67 & & 20.5 & 5.09 + 4489.74 & & 0.12 & @xmath1333.90 & 60.0 & 4.60 & & 73.8 & 4.79 & & & + 4494.56 & & 2.20 & @xmath1331.14 & 60.0 & 4.37 & & 78.8 & 4.65 & & 79.6 & 4.83 + 4528.61 & & 2.18 & @xmath1330.82 & 77.5 & 4.32 & & & & & 96.9 & 4.82 + 4531.15 & & 1.48 & @xmath1332.10 & 58.7 & 4.43 & & 70.0 & 4.59 & & 68.5 & 4.71 + 4592.65 & & 1.56 & @xmath1332.46 & 43.4 & 4.63 & & 53.9 & 4.79 & & 55.8 & 4.94 + 4602.94 & & 1.49 & @xmath1332.21 & 58.7 & 4.53 & & 67.7 & 4.65 & & 65.9 & 4.77 + 4630.12 & & 2.28 & @xmath1332.59 & & & & 10.4 & 4.84 & & 13.2 & 5.03 + 4632.91 & & 1.61 & @xmath1332.91 & 18.1 & 4.62 & & 19.9 & 4.67 & & 21.4 & 4.81 + 4647.43 & & 2.95 & @xmath1331.35 & 14.8 & 4.55 & & 19.3 & 4.70 & & 18.8 & 4.76 + 4678.85 & & 3.60 & @xmath1330.83 & & & & 16.3 & 4.85 & & 19.1 & 5.00 + 4691.41 & & 2.99 & @xmath1331.52 & & & & 18.1 & 4.87 & & 20.0 & 5.00 + 4707.27 & & 3.24 & @xmath1331.08 & 16.0 & 4.66 & & 25.1 & 4.90 & & 23.7 & 4.94 + 4710.28 & & 3.02 & @xmath1331.61 & 9.4 & 4.67 & & 19.0 & 5.02 & & 14.7 & 4.96 + 4733.59 & & 1.49 & @xmath1332.99 & 20.1 & 4.59 & & 28.4 & 4.78 & & 27.5 & 4.87 + 4736.77 & & 3.21 & @xmath1330.75 & 23.5 & 4.49 & & 35.0 & 4.72 & & 39.8 & 4.89 + 4786.81 & & 3.00 & @xmath1331.61 & & & & & & & 13.7 & 4.89 + 4789.65 & & 3.53 & @xmath1330.96 & & & & & & & 9.8 & 4.70 + 4859.74 & & 2.88 & @xmath1330.76 & 52.2 & 4.61 & & 56.6 & 4.67 & & 62.1 & 4.87 + 4871.32 & & 2.87 & @xmath1330.36 & 59.1 & 4.30 & & 69.5 & 4.44 & & 70.8 & 4.59 + 4872.14 & & 2.88 & @xmath1330.57 & 47.3 & 4.34 & & 57.7 & 4.49 & & 66.2 & 4.74 + 4890.76 & & 2.88 & @xmath1330.39 & 54.9 & 4.28 & & 66.3 & 4.44 & & 72.7 & 4.67 + 4891.49 & & 2.85 & @xmath1330.11 & 72.1 & 4.23 & & 81.1 & 4.35 & & 85.7 & 4.58 + 4903.31 & & 2.88 & @xmath1330.93 & 29.6 & 4.39 & & 34.8 & 4.49 & & 44.9 & 4.75 + 4918.99 & & 2.85 & @xmath1330.34 & 63.5 & 4.32 & & & & & 71.3 & 4.54 + 4920.50 & & 2.83 & + 0.07 & 85.2 & 4.23 & & 92.3 & 4.31 & & 96.3 & 4.56 + 4924.77 & & 2.28 & @xmath1332.11 & & & & 23.5 & 4.72 & & 19.4 & 4.71 + 4938.81 & & 2.88 & @xmath1331.08 & 26.8 & 4.48 & & 36.8 & 4.67 & & 35.9 & 4.73 + 4939.69 & & 0.86 & @xmath1333.25 & 40.9 & 4.46 & & 52.7 & 4.65 & & 49.5 & 4.73 + 4946.39 & & 3.37 & @xmath1331.17 & & & & 15.8 & 4.87 & & 14.5 & 4.89 + 4966.09 & & 3.33 & @xmath1330.87 & 17.1 & 4.55 & & 31.5 & 4.89 & & 25.8 & 4.85 + 4973.10 & & 3.96 & @xmath1330.95 & & & & 12.8 & 5.24 & & & + 4994.13 & & 0.92 & @xmath1332.97 & 62.0 & 4.57 & & 62.8 & 4.57 & & 55.3 & 4.60 + 5001.87 & & 3.88 & + 0.05 & 22.8 & 4.43 & & 22.0 & 4.42 & & 23.3 & 4.51 + 5006.12 & & 2.83 & @xmath1330.61 & 48.6 & 4.32 & & 58.1 & 4.46 & & & + 5012.07 & & 0.86 & @xmath1332.64 & 75.6 & 4.36 & & 97.2 & 4.67 & & 94.4 & 4.84 + 5022.24 & & 3.98 & @xmath1330.53 & & & & & & & 8.7 & 4.70 + 5041.07 & & 0.96 & @xmath1333.09 & 62.8 & 4.74 & & 71.1 & 4.85 & & 67.8 & 4.96 + 5041.76 & & 1.49 & @xmath1332.20 & 63.4 & 4.51 & & 81.3 & 4.75 & & 83.4 & 4.96 + 5049.82 & & 2.28 & @xmath1331.35 & 45.9 & 4.35 & & 66.2 & 4.65 & & 48.8 & 4.50 + 5051.63 & & 0.92 & @xmath1332.76 & 73.8 & 4.52 & & 88.4 & 4.71 & & 76.3 & 4.71 + 5060.08 & & 0.00 & @xmath1335.46 & & & & 17.4 & 5.14 & & & + 5068.77 & & 2.94 & @xmath1331.04 & 26.4 & 4.48 & & 30.2 & 4.56 & & 30.4 & 4.65 + 5074.75 & & 4.22 & @xmath1330.20 & & & & 19.4 & 4.99 & & 16.5 & 4.96 + 5079.22 & & 2.20 & @xmath1332.10 & 23.4 & 4.59 & & 27.4 & 4.68 & & 31.3 & 4.85 + 5079.74 & & 0.99 & @xmath1333.25 & 33.3 & 4.47 & & 56.6 & 4.84 & & 50.2 & 4.88 + 5083.34 & & 0.96 & @xmath1332.84 & 55.6 & 4.38 & & 68.7 & 4.56 & & 68.7 & 4.71 + 5090.77 & & 4.26 & @xmath1330.36 & & & & & & & 9.6 & 4.90 + 5098.70 & & 2.18 & @xmath1332.03 & 37.1 & 4.76 & & & & & 39.8 & 4.91 + 5110.41 & & 0.00 & @xmath1333.76 & 87.0 & 4.57 & & 95.0 & 4.67 & & 95.9 & 4.91 + 5123.72 & & 1.01 & @xmath1333.06 & 57.9 & 4.68 & & 63.5 & 4.76 & & 64.3 & 4.92 + 5125.12 & & 4.22 & @xmath1330.14 & & & & 13.7 & 4.75 & & 22.0 & 5.05 + 5127.36 & & 0.92 & @xmath1333.25 & 48.4 & 4.62 & & 58.1 & 4.76 & & 54.2 & 4.85 + 5131.47 & & 2.22 & @xmath1332.52 & 9.7 & 4.58 & & 11.1 & 4.66 & & 26.3 & 5.19 + 5133.69 & & 4.18 & + 0.14 & & & & 30.1 & 4.84 & & 33.1 & 4.96 + 5137.38 & & 4.18 & @xmath1330.40 & & & & 9.1 & 4.77 & & 12.8 & 4.98 + 5141.74 & & 2.42 & @xmath1332.24 & 13.2 & 4.69 & & & & & 18.0 & 4.94 + 5142.93 & & 0.96 & @xmath1333.08 & 54.0 & 4.59 & & 65.1 & 4.74 & & 62.5 & 4.85 + 5150.84 & & 0.99 & @xmath1333.04 & 41.2 & 4.39 & & 55.7 & 4.60 & & 55.4 & 4.74 + 5151.91 & & 1.01 & @xmath1333.32 & 28.4 & 4.47 & & & & & 41.8 & 4.83 + 5162.27 & & 4.18 & + 0.02 & 20.8 & 4.75 & & 27.9 & 4.91 & & 25.9 & 4.93 + 5166.28 & & 0.00 & @xmath1334.12 & 60.0 & 4.53 & & 69.0 & 4.65 & & 71.8 & 4.86 + 5171.60 & & 1.49 & @xmath1331.72 & & & & 99.7 & 4.53 & & 94.7 & 4.65 + 5191.45 & & 3.04 & @xmath1330.55 & 38.3 & 4.32 & & 48.5 & 4.48 & & 53.5 & 4.66 + 5192.34 & & 3.00 & @xmath1330.42 & 51.8 & 4.36 & & 53.3 & 4.38 & & 61.2 & 4.61 + 5194.94 & & 1.56 & @xmath1332.02 & 68.1 & 4.45 & & 78.6 & 4.59 & & 73.1 & 4.66 + 5198.71 & & 2.22 & @xmath1332.09 & 18.1 & 4.45 & & & & & 24.5 & 4.71 + 5202.34 & & 2.18 & @xmath1331.87 & 31.8 & 4.49 & & 50.9 & 4.80 & & 47.2 & 4.85 + 5216.27 & & 1.61 & @xmath1332.08 & 54.8 & 4.38 & & 71.1 & 4.60 & & 77.2 & 4.84 + 5217.39 & & 3.21 & @xmath1331.16 & & & & 22.7 & 4.83 & & 21.9 & 4.88 + 5225.53 & & 0.11 & @xmath1334.75 & 26.5 & 4.75 & & 31.5 & 4.86 & & 23.0 & 4.82 + 5232.94 & & 2.94 & @xmath1330.06 & 75.6 & 4.28 & & 79.4 & 4.31 & & 82.6 & 4.52 + 5242.49 & & 3.63 & @xmath1330.97 & & & & & & & 10.7 & 4.82 + 5247.05 & & 0.09 & @xmath1334.95 & 14.5 & 4.61 & & 23.2 & 4.86 & & 14.8 & 4.76 + 5250.21 & & 0.12 & @xmath1334.94 & 19.5 & 4.79 & & 23.6 & 4.90 & & 14.8 & 4.78 + 5250.65 & & 2.20 & @xmath1332.18 & 27.5 & 4.74 & & 36.3 & 4.90 & & 39.4 & 5.05 + 5254.96 & & 0.11 & @xmath1334.76 & & & & 43.8 & 5.06 & & 31.1 & 4.99 + 5263.31 & & 3.27 & @xmath1330.88 & & & & 26.9 & 4.71 & & 26.6 & 4.78 + 5266.56 & & 3.00 & @xmath1330.39 & 51.7 & 4.32 & & 65.7 & 4.52 & & 65.6 & 4.64 + 5269.54 & & 0.86 & @xmath1331.33 & 142.6 & 4.17 & & 152.4 & 4.25 & & 153.9 & 4.59 + 5281.79 & & 3.04 & @xmath1330.83 & 28.1 & 4.40 & & 35.8 & 4.55 & & 39.5 & 4.70 + 5283.62 & & 3.24 & @xmath1330.52 & 38.9 & 4.53 & & & & & 45.7 & 4.73 + 5302.30 & & 3.28 & @xmath1330.72 & 19.2 & 4.37 & & 30.8 & 4.63 & & 31.3 & 4.72 + 5307.36 & & 1.61 & @xmath1332.91 & 16.2 & 4.47 & & 32.2 & 4.85 & & 21.0 & 4.71 + 5324.18 & & 3.21 & @xmath1330.10 & 52.9 & 4.29 & & 63.9 & 4.44 & & 57.8 & 4.46 + 5328.04 & & 0.92 & @xmath1331.47 & 133.8 & 4.20 & & 146.1 & 4.33 & & 142.2 & 4.56 + 5328.53 & & 1.56 & @xmath1331.85 & 80.3 & 4.44 & & 91.7 & 4.59 & & 92.5 & 4.79 + 5332.90 & & 1.55 & @xmath1332.78 & & & & 37.7 & 4.73 & & 32.5 & 4.75 + 5339.93 & & 3.27 & @xmath1330.72 & 24.6 & 4.49 & & 37.5 & 4.74 & & 37.5 & 4.82 + 5364.87 & & 4.45 & + 0.23 & & & & 7.6 & 4.35 & & 20.3 & 4.88 + 5367.47 & & 4.42 & + 0.44 & 22.1 & 4.63 & & 21.1 & 4.61 & & 25.6 & 4.77 + 5369.96 & & 4.37 & + 0.54 & 16.7 & 4.33 & & & & & 25.4 & 4.60 + 5371.49 & & 0.96 & @xmath1331.64 & 133.0 & 4.39 & & 137.7 & 4.39 & & & + 5383.37 & & 4.31 & + 0.65 & 30.5 & 4.47 & & 35.3 & 4.56 & & 39.2 & 4.70 + 5393.17 & & 3.24 & @xmath1330.91 & 26.0 & 4.67 & & 32.0 & 4.79 & & 34.6 & 4.91 + 5397.13 & & 0.92 & @xmath1331.98 & 114.0 & 4.32 & & 126.4 & 4.48 & & 122.9 & 4.69 + 5405.77 & & 0.99 & @xmath1331.85 & 118.6 & 4.36 & & 121.3 & 4.34 & & 122.4 & 4.63 + 5410.91 & & 4.47 & + 0.40 & 17.6 & 4.60 & & 21.8 & 4.72 & & & + 5415.20 & & 4.39 & + 0.64 & 22.6 & 4.40 & & 26.4 & 4.49 & & 32.8 & 4.68 + 5424.07 & & 4.32 & + 0.52 & 31.2 & 4.62 & & 30.9 & 4.62 & & 38.6 & 4.82 + 5429.70 & & 0.96 & @xmath1331.88 & 116.8 & 4.31 & & 128.0 & 4.45 & & 123.2 & 4.64 + 5434.52 & & 1.01 & @xmath1332.13 & 103.6 & 4.40 & & 109.7 & 4.46 & & 117.8 & 4.84 + 5446.92 & & 0.99 & @xmath1331.91 & 112.5 & 4.30 & & 130.0 & 4.54 & & 131.6 & 4.85 + 5455.61 & & 1.01 & @xmath1332.09 & 119.4 & 4.63 & & 117.0 & 4.52 & & 131.1 & 5.04 + 5497.52 & & 1.01 & @xmath1332.83 & 63.7 & 4.49 & & & & & 76.1 & 4.81 + 5501.47 & & 0.96 & @xmath1333.05 & 59.3 & 4.58 & & 70.2 & 4.73 & & 68.2 & 4.85 + 5506.78 & & 0.99 & @xmath1332.79 & 72.4 & 4.54 & & 81.6 & 4.66 & & & + 5569.62 & & 3.42 & @xmath1330.54 & 24.3 & 4.46 & & 37.9 & 4.72 & & 34.0 & 4.73 + 5572.84 & & 3.40 & @xmath1330.28 & 38.9 & 4.46 & & 42.5 & 4.51 & & 44.0 & 4.62 + 5576.09 & & 3.43 & @xmath1331.00 & 13.1 & 4.62 & & & & & 24.1 & 5.00 + 5586.76 & & 3.37 & @xmath1330.14 & 40.7 & 4.31 & & 52.3 & 4.48 & & 56.7 & 4.64 + 5615.64 & & 3.33 & + 0.05 & 57.8 & 4.33 & & 63.8 & 4.40 & & 68.1 & 4.58 + 5624.54 & & 3.42 & @xmath1330.76 & 18.9 & 4.54 & & 18.2 & 4.53 & & 25.0 & 4.77 + 5658.82 & & 3.40 & @xmath1330.79 & 16.0 & 4.46 & & 27.7 & 4.76 & & 27.1 & 4.82 + 5701.54 & & 2.56 & @xmath1332.14 & & & & 13.2 & 4.72 & & 13.9 & 4.82 + 5753.12 & & 4.26 & @xmath1330.69 & & & & & & & 10.1 & 5.20 + 6065.48 & & 2.61 & @xmath1331.41 & 32.5 & 4.49 & & 34.6 & 4.53 & & 37.7 & 4.67 + 6136.61 & & 2.45 & @xmath1331.41 & 40.8 & 4.43 & & 54.4 & 4.63 & & 50.7 & 4.69 + 6137.69 & & 2.59 & @xmath1331.35 & 30.8 & 4.36 & & 44.5 & 4.59 & & & + 6191.56 & & 2.43 & @xmath1331.42 & 36.6 & 4.34 & & 53.2 & 4.59 & & 52.8 & 4.70 + 6213.43 & & 2.22 & @xmath1332.48 & & & & & & & 20.2 & 4.91 + 6219.28 & & 2.20 & @xmath1332.45 & 16.7 & 4.66 & & 25.2 & 4.88 & & 26.1 & 5.00 + 6230.72 & & 2.56 & @xmath1331.28 & 42.4 & 4.45 & & 55.3 & 4.64 & & 54.7 & 4.74 + 6246.32 & & 3.60 & @xmath1330.88 & 14.1 & 4.68 & & 8.4 & 4.45 & & 17.3 & 4.86 + 6252.56 & & 2.40 & @xmath1331.69 & 33.2 & 4.51 & & 38.7 & 4.61 & & 41.8 & 4.76 + 6254.26 & & 2.28 & @xmath1332.44 & 17.1 & 4.76 & & 17.1 & 4.76 & & 19.5 & 4.92 + 6265.13 & & 2.18 & @xmath1332.54 & 14.7 & 4.66 & & 17.9 & 4.77 & & 15.3 & 4.78 + 6297.79 & & 2.22 & @xmath1332.64 & & & & 33.8 & 5.26 & & & + 6301.50 & & 3.65 & @xmath1330.72 & & & & 16.1 & 4.65 & & 24.5 & 4.94 + 6322.68 & & 2.59 & @xmath1332.47 & & & & & & & 13.7 & 5.14 + 6335.33 & & 2.20 & @xmath1332.18 & 25.2 & 4.60 & & 22.9 & 4.55 & & 25.0 & 4.70 + 6336.84 & & 3.69 & @xmath1331.05 & & & & 10.5 & 4.82 & & 11.2 & 4.92 + 6355.03 & & 2.84 & @xmath1332.29 & & & & & & & 9.7 & 5.09 + 6393.60 & & 2.43 & @xmath1331.58 & 35.9 & 4.47 & & 52.4 & 4.72 & & 46.0 & 4.74 + 6400.00 & & 3.60 & @xmath1330.29 & 30.1 & 4.49 & & 27.8 & 4.45 & & 38.3 & 4.71 + 6411.65 & & 3.65 & @xmath1330.59 & 13.1 & 4.41 & & 10.9 & 4.33 & & 22.4 & 4.76 + 6421.35 & & 2.28 & @xmath1332.01 & 24.2 & 4.49 & & 33.1 & 4.67 & & 35.4 & 4.81 + 6430.85 & & 2.18 & @xmath1331.95 & 31.8 & 4.47 & & 41.9 & 4.64 & & 40.3 & 4.72 + 6494.98 & & 2.40 & @xmath1331.24 & 55.4 & 4.39 & & 74.0 & 4.63 & & & + 6498.94 & & 0.96 & @xmath1334.70 & & & & & & & 8.9 & 5.21 + 6592.91 & & 2.73 & @xmath1331.47 & 20.4 & 4.40 & & 33.8 & 4.68 & & 27.4 & 4.65 + 6593.87 & & 2.44 & @xmath1332.37 & 9.1 & 4.55 & & 11.0 & 4.65 & & 16.6 & 4.94 + 6663.44 & & 2.42 & @xmath1332.48 & 12.6 & 4.79 & & 13.3 & 4.82 & & 14.4 & 4.95 + 6677.99 & & 2.69 & @xmath1331.42 & 24.8 & 4.40 & & 38.9 & 4.66 & & 40.8 & 4.79 + 6750.15 & & 2.42 & @xmath1332.58 & & & & 8.6 & 4.71 & & & + 6978.85 & & 2.48 & @xmath1332.45 & & & & 8.7 & 4.65 & & 20.1 & 5.14 + 4122.67 & & 2.58 & @xmath1333.38 & & & & 35.6 & 5.23 & & 40.7 & 5.32 + 4128.75 & & 2.58 & @xmath1333.47 & & & & & & & 10.5 & 4.64 + 4178.86 & & 2.58 & @xmath1332.51 & 36.0 & 4.37 & & 51.2 & 4.62 & & 62.4 & 4.82 + 4233.17 & & 2.58 & @xmath1331.97 & 71.8 & 4.43 & & 77.9 & 4.50 & & 89.4 & 4.78 + 4416.82 & & 2.78 & @xmath1332.60 & 30.3 & 4.54 & & 31.2 & 4.54 & & 37.6 & 4.65 + 4489.19 & & 2.83 & @xmath1332.97 & 14.6 & 4.56 & & 15.1 & 4.56 & & 19.6 & 4.68 + 4491.41 & & 2.86 & @xmath1332.71 & 19.3 & 4.48 & & 23.1 & 4.56 & & 30.4 & 4.70 + 4508.28 & & 2.86 & @xmath1332.58 & 33.7 & 4.66 & & 40.3 & 4.76 & & 50.0 & 4.92 + 4515.34 & & 2.84 & @xmath1332.60 & 32.3 & 4.63 & & 38.5 & 4.72 & & 43.9 & 4.81 + 4520.22 & & 2.81 & @xmath1332.60 & 31.5 & 4.58 & & 40.2 & 4.72 & & 40.7 & 4.72 + 4522.63 & & 2.84 & @xmath1332.25 & 53.2 & 4.63 & & 57.6 & 4.68 & & 78.1 & 5.05 + 4541.52 & & 2.86 & @xmath1333.05 & & & & 23.4 & 4.90 & & 23.9 & 4.89 + 4555.89 & & 2.83 & @xmath1332.40 & 33.9 & 4.44 & & 48.6 & 4.67 & & 53.3 & 4.75 + 4576.34 & & 2.84 & @xmath1332.95 & & & & 19.3 & 4.67 & & 21.7 & 4.71 + 4583.84 & & 2.81 & @xmath1331.93 & 66.0 & 4.46 & & 71.4 & 4.51 & & 83.5 & 4.77 + 4620.52 & & 2.83 & @xmath1333.21 & & & & 13.8 & 4.74 & & 13.6 & 4.71 + 4731.44 & & 2.89 & @xmath1333.36 & & & & & & & 19.3 & 5.10 + 4923.93 & & 2.89 & @xmath1331.32 & 85.8 & 4.19 & & 88.0 & 4.18 & & 105.5 & 4.55 + 5018.45 & & 2.89 & @xmath1331.22 & 100.5 & 4.32 & & 99.3 & 4.23 & & 122.6 & 4.76 + 5197.58 & & 3.23 & @xmath1332.22 & 29.6 & 4.57 & & 28.5 & 4.53 & & 39.5 & 4.72 + 5234.63 & & 3.22 & @xmath1332.18 & 26.9 & 4.46 & & 32.0 & 4.54 & & 43.7 & 4.74 + 5276.00 & & 3.20 & @xmath1332.01 & 32.9 & 4.38 & & 39.8 & 4.48 & & 53.4 & 4.69 + 5284.08 & & 2.89 & @xmath1333.19 & & & & & & & 18.8 & 4.85 + 5534.83 & & 3.25 & @xmath1332.93 & 5.6 & 4.44 & & 22.6 & 5.11 & & 14.5 & 4.86 + 6247.55 & & 3.89 & @xmath1332.51 & & & & 9.5 & 4.97 & & 8.4 & 4.88 + 6456.38 & & 3.90 & @xmath1332.08 & & & & & & & 14.8 & 4.73 + 3845.47 & & 0.92 & + 0.01 & & 1.72 & & & 1.54 & & & 2.24 + 3873.12 & & 0.43 & @xmath1330.66 & & 1.97 & & & 2.38 & & & 2.39 + 3881.87 & & 0.58 & @xmath1331.13 & & 1.92 & & & 2.04 & & & 2.29 + 3995.31 & & 0.92 & @xmath1330.22 & & 1.57 & & & 1.79 & & & 1.89 + 4020.90 & & 0.43 & @xmath1332.07 & & 1.92 & & & 2.19 & & & 2.29 + 4110.53 & & 1.05 & @xmath1331.08 & & 1.72 & & & 1.69 & & & 1.99 + 4118.77 & & 1.05 & @xmath1330.49 & & 1.72 & & & 1.94 & & & 2.09 + 4121.32 & & 0.92 & @xmath1330.32 & & 1.72 & & & 1.89 & & & 1.99 + 3597.71 & & 0.21 & @xmath1331.11 & 103.6 & 3.35 & & & & & & + 3783.52 & & 0.42 & @xmath1331.42 & 104.3 & 3.55 & & 124.6 & 3.91 & & & + 3807.14 & & 0.42 & @xmath1331.22 & 91.8 & 3.03 & & 109.0 & 3.33 & & 113.7 & 3.68 + 3858.30 & & 0.42 & @xmath1330.95 & 117.3 & 3.30 & & 115.8 & 3.15 & & 122.2 & 3.55 + 4648.66 & & 3.42 & @xmath1330.16 & & & & 13.1 & 3.57 & & 15.7 & 3.72 + 4714.42 & & 3.38 & + 0.23 & 14.7 & 3.18 & & 28.1 & 3.52 & & 33.1 & 3.69 + 4855.41 & & 3.54 & + 0.00 & & & & & & & 13.7 & 3.60 + 4904.41 & & 3.54 & @xmath1330.17 & & & & & & & 12.0 & 3.70 + 4980.16 & & 3.61 & @xmath1330.11 & & & & 7.8 & 3.45 & & 19.7 & 3.97 + 5017.59 & & 3.54 & @xmath1330.08 & & & & 20.4 & 3.81 & & & + 5035.37 & & 3.63 & + 0.29 & & & & & & & 18.9 & 3.56 + 5080.52 & & 3.65 & + 0.13 & & & & 20.2 & 3.72 & & 24.0 & 3.87 + 5081.11 & & 3.85 & + 0.30 & & & & 17.1 & 3.69 & & 14.3 & 3.66 + 5084.08 & & 3.68 & + 0.03 & & & & 16.2 & 3.74 & & 11.4 & 3.62 + 5137.07 & & 1.68 & @xmath1331.99 & & & & 28.0 & 3.66 & & 32.4 & 3.84 + 5476.90 & & 1.83 & @xmath1330.89 & & & & 69.9 & 3.34 & & 77.5 & 3.59 + 5578.73 & & 1.68 & @xmath1332.64 & & & & & & & 6.6 & 3.64 + 5754.68 & & 1.94 & @xmath1332.33 & & & & & & & 11.0 & 3.86 + 6643.64 & & 1.68 & @xmath1332.30 & 10.3 & 3.33 & & & & & 18.7 & 3.72 + 6767.77 & & 1.83 & @xmath1332.17 & 8.5 & 3.28 & & 13.5 & 3.51 & & 20.3 & 3.81 + 5105.54 & & 1.39 & @xmath1331.52 & & 0.92 & & & : 0.84 & & & 0.99 + 4722.15 & & 4.03 & @xmath1330.39 & & & & 12.7 & 2.00 & & 17.4 & 2.19 + 4810.53 & & 4.08 & @xmath1330.14 & 14.5 & 1.87 & & 19.2 & 2.01 & & 19.2 & 2.04 + 4077.71 & & 0.00 & + 0.15 & 149.9 & @xmath1330.50 & & 159.9 & @xmath1330.44 & & 155.3 & @xmath1330.30 + 4215.52 & & 0.00 & @xmath1330.18 & 137.2 & @xmath1330.52 & & 151.5 & @xmath1330.37 & & 138.7 & @xmath1330.39 + 4883.69 & & 1.08 & + 0.07 & 12.6 & @xmath1331.19 & & 20.0 & @xmath1330.96 & & 20.2 & @xmath1330.94 + 5087.43 & & 1.08 & @xmath1330.17 & 8.4 & @xmath1331.17 & & 12.8 & @xmath1330.97 & & 13.4 & @xmath1330.93 + 4554.03 & & 0.00 & + 0.16 & & @xmath1331.44 & & & @xmath1331.82 & & & @xmath1331.02 + 4934.09 & & 0.00 & @xmath1330.16 & & @xmath1331.24 & & & @xmath1331.91 & & & @xmath1331.12 + 4809.0 & & 0.24 & & & @xmath1330.92 & & & @xmath1330.65 & & & @xmath1330.80 + 6645.1 & & 1.38 & 0.20 & & @xmath1332.00 & & & @xmath1331.83 & & & @xmath1331.68 lcrrrr & + + & 1 & + 0.25 & @xmath1330.80 & + 2.22 & + c ( ch ) & 1 & + 5.41 & @xmath1333.02 & + 0.00 & + & 2 & + 3.40 & @xmath1332.85 & + 0.18 & @xmath1750.01 + & 10 & + 5.12 & @xmath1332.48 & + 0.54 & 0.08 + & 1 & + 2.73 & @xmath1333.72 & @xmath1330.70 & + & 1 & + 5.29 & @xmath1332.22 & + 0.81 & + & 2 & + 2.66 & @xmath1332.38 & + 0.65 & 0.11 + & 18 & + 3.73 & @xmath1332.61 & + 0.41 & 0.10 + & 6 & @xmath1330.46 & @xmath1333.61 & @xmath1330.59 & 0.14 + & 24 & + 2.25 & @xmath1332.70 & + 0.32 & 0.15 + & 52 & + 2.19 & @xmath1332.76 & + 0.26 & 0.19 + & 1 & + 0.46 & @xmath1333.47 & @xmath1330.45 & + & 12 & + 2.35 & @xmath1333.28 & @xmath1330.26 & 0.23 + & 2 & + 2.79 & @xmath1332.85 & + 0.17 & 0.04 + & 4 & + 1.50 & @xmath1333.93 & @xmath1330.91 & 0.11 + & 201 & + 4.48 & @xmath1333.02 & + 0.00 & 0.14 + & 17 & + 4.48 & @xmath1333.02 & + 0.00 & 0.12 + & 8 & + 1.78 & @xmath1333.21 & @xmath1330.19 & 0.13 + & 7 & + 3.29 & @xmath1332.93 & + 0.09 & 0.15 + & 1 & @xmath175 + 0.92 & @xmath1763.27 & @xmath1760.25 & + & 1 & + 1.87 & @xmath1332.69 & + 0.33 & + & 2 & @xmath1330.51 & @xmath1333.38 & @xmath1330.36 & 0.01 + & 2 & @xmath1331.18 & @xmath1333.39 & @xmath1330.37 & 0.01 + & 2 & @xmath1331.34 & @xmath1333.52 & @xmath1330.50 & 0.10 + & 1 & @xmath1760.92 & @xmath1762.02 & @xmath1771.00 & + & 1 & @xmath1762.00 & @xmath1762.52 & @xmath1770.50 & + + + & 1 & + 0.40 & @xmath1330.65 & + 2.19 & + c ( ch ) & 1 & + 5.74 & @xmath1332.69 & + 0.15 & + & 2 & + 3.46 & @xmath1332.78 & + 0.07 & 0.05 + & 9 & + 5.33 & @xmath1332.27 & + 0.57 & 0.13 + & 1 & + 3.07 & @xmath1333.38 & @xmath1330.54 & + & 1 & + 5.52 & @xmath1331.98 & + 0.86 & + & 2 & + 2.90 & @xmath1332.13 & + 0.71 & 0.13 + & 23 & + 3.96 & @xmath1332.38 & + 0.47 & 0.13 + & 7 & @xmath1330.23 & @xmath1333.38 & @xmath1330.54 & 0.15 + & 23 & + 2.38 & @xmath1332.57 & + 0.28 & 0.10 + & 50 & + 2.31 & @xmath1332.64 & + 0.20 & 0.18 + & 1 & + 0.58 & @xmath1333.35 & @xmath1330.51 & + & 18 & + 2.55 & @xmath1333.09 & @xmath1330.25 & 0.25 + & 2 & + 2.98 & @xmath1332.66 & + 0.18 & 0.04 + & 4 & + 1.71 & @xmath1333.71 & @xmath1330.87 & 0.10 + & 216 & + 4.66 & @xmath1332.84 & + 0.00 & 0.17 + & 22 & + 4.66 & @xmath1332.84 & + 0.00 & 0.24 + & 8 & + 1.93 & @xmath1333.06 & @xmath1330.21 & 0.25 + & 13 & + 3.57 & @xmath1332.65 & + 0.19 & 0.21 + & 1 & : + 0.84 & : @xmath1333.35 & : @xmath1330.51 & + & 2 & + 2.00 & @xmath1332.55 & + 0.29 & @xmath1750.01 + & 2 & @xmath1330.41 & @xmath1333.28 & @xmath1330.43 & 0.04 + & 2 & @xmath1330.96 & @xmath1333.17 & @xmath1330.33 & 0.01 + & 2 & @xmath1331.86 & @xmath1334.04 & @xmath1331.20 & 0.05 + & 1 & @xmath1760.65 & @xmath1761.75 & @xmath1771.09 & + & 1 & @xmath1761.83 & @xmath1762.35 & @xmath1770.49 & + + + & 1 & + 0.10 & @xmath1330.95 & + 1.75 & + c ( ch ) & 1 & + 5.12 & @xmath1333.31 & @xmath1330.61 & + & 2 & + 3.92 & @xmath1332.33 & + 0.37 & 0.04 + & 10 & + 5.36 & @xmath1332.24 & + 0.46 & 0.08 + & 1 & + 3.27 & @xmath1333.18 & @xmath1330.48 & + & 1 & + 5.52 & @xmath1331.99 & + 0.71 & + & 1 & + 2.94 & @xmath1332.09 & + 0.61 & + & 23 & + 4.06 & @xmath1332.28 & + 0.42 & 0.14 + & 8 & @xmath1330.11 & @xmath1333.26 & @xmath1330.56 & 0.11 + & 25 & + 2.49 & @xmath1332.46 & + 0.24 & 0.11 + & 51 & + 2.46 & @xmath1332.49 & + 0.21 & 0.17 + & 1 & + 0.68 & @xmath1333.25 & @xmath1330.55 & + & 16 & + 2.68 & @xmath1332.96 & @xmath1330.26 & 0.11 + & 2 & + 3.10 & @xmath1332.53 & + 0.16 & 0.02 + & 4 & + 1.94 & @xmath1333.49 & @xmath1330.79 & 0.14 + & 225 & + 4.80 & @xmath1332.70 & + 0.00 & 0.14 + & 26 & + 4.80 & @xmath1332.70 & @xmath1330.00 & 0.16 + & 8 & + 2.15 & @xmath1332.84 & @xmath1330.15 & 0.17 + & 17 & + 3.71 & @xmath1332.51 & + 0.19 & 0.12 + & 1 & @xmath175 + 0.99 & @xmath1763.20 & @xmath1760.50 & + & 2 & + 2.12 & @xmath1332.44 & + 0.25 & 0.07 + & 2 & @xmath1330.34 & @xmath1333.21 & @xmath1330.52 & 0.05 + & 2 & @xmath1330.94 & @xmath1333.15 & @xmath1330.45 & @xmath1750.01 + & 2 & @xmath1331.07 & @xmath1333.25 & @xmath1330.55 & 0.05 + & 1 & @xmath1760.80 & @xmath1761.90 & @xmath1770.80 & + & 1 & @xmath1761.68 & @xmath1762.20 & @xmath1770.50 & lcccccccc & + + na i & + 0.15 & @xmath1330.02 & @xmath1330.04 & @xmath1330.01 & & 0.07 & & 0.17 + mg i & + 0.08 & @xmath1330.08 & @xmath1330.02 & @xmath1330.01 & & 0.03 & & 0.12 + al i & + 0.04 & @xmath1330.11 & @xmath1330.05 & @xmath1330.02 & & 0.10 & & 0.16 + si i & + 0.08 & @xmath1330.09 & @xmath1330.03 & @xmath1330.01 & & 0.07 & & 0.14 + k i & + 0.09 & @xmath1330.01 & @xmath1330.01 & + 0.00 & & 0.08 & & 0.12 + ca i & + 0.06 & @xmath1330.03 & @xmath1330.01 & @xmath1330.00 & & 0.02 & & 0.07 + sc ii & + 0.06 & + 0.04 & @xmath1330.01 & + 0.00 & & 0.04 & & 0.08 + ti i & + 0.10 & @xmath1330.04 & @xmath1330.01 & @xmath1330.00 & & 0.03 & & 0.11 + ti ii & + 0.04 & + 0.03 & @xmath1330.02 & @xmath1330.00 & & 0.03 & & 0.06 + v i & + 0.09 & @xmath1330.04 & @xmath1330.01 & @xmath1330.01 & & 0.10 & & 0.14 + v ii & + 0.02 & + 0.03 & @xmath1330.00 & + 0.00 & & 0.06 & & 0.07 + cr i & + 0.10 & @xmath1330.04 & @xmath1330.02 & @xmath1330.00 & & 0.07 & & 0.12 + cr ii & @xmath1330.01 & + 0.05 & + 0.00 & + 0.00 & & 0.07 & & 0.09 + mn i & + 0.08 & @xmath1330.08 & @xmath1330.04 & @xmath1330.01 & & 0.07 & & 0.14 + fe i & + 0.08 & @xmath1330.04 & @xmath1330.01 & @xmath1330.00 & & 0.01 & & 0.09 + fe ii & + 0.01 & + 0.05 & @xmath1330.01 & + 0.00 & & 0.03 & & 0.06 + co i & + 0.09 & @xmath1330.07 & @xmath1330.02 & @xmath1330.01 & & 0.06 & & 0.13 + ni i & + 0.07 & @xmath1330.08 & @xmath1330.04 & @xmath1330.01 & & 0.06 & & 0.13 + zn i & + 0.04 & + 0.01 & + 0.00 & + 0.00 & & 0.10 & & 0.11 + sr ii & + 0.08 & + 0.01 & @xmath1330.07 & @xmath1330.01 & & 0.07 & & 0.13 + y ii & + 0.05 & + 0.05 & + 0.00 & + 0.00 & & 0.07 & & 0.10 + ba ii & + 0.10 & + 0.04 & @xmath1330.03 & @xmath1330.00 & & 0.06 & & 0.13 + + + na i & + 0.14 & @xmath1330.04 & @xmath1330.04 & @xmath1330.01 & & 0.07 & & 0.17 + mg i & + 0.06 & @xmath1330.07 & @xmath1330.02 & @xmath1330.01 & & 0.04 & & 0.11 + al i & + 0.04 & @xmath1330.14 & @xmath1330.06 & @xmath1330.03 & & 0.10 & & 0.19 + si i & + 0.09 & @xmath1330.11 & @xmath1330.03 & @xmath1330.01 & & 0.07 & & 0.16 + k i & + 0.09 & @xmath1330.01 & @xmath1330.01 & + 0.00 & & 0.09 & & 0.13 + ca i & + 0.06 & @xmath1330.03 & @xmath1330.01 & @xmath1330.00 & & 0.03 & & 0.07 + sc ii & + 0.05 & + 0.03 & @xmath1330.02 & @xmath1330.00 & & 0.07 & & 0.10 + ti i & + 0.10 & @xmath1330.03 & @xmath1330.01 & @xmath1330.00 & & 0.02 & & 0.11 + ti ii & + 0.04 & + 0.03 & @xmath1330.02 & @xmath1330.00 & & 0.03 & & 0.06 + v i & + 0.10 & @xmath1330.03 & + 0.01 & + 0.00 & & 0.10 & & 0.14 + v ii & + 0.02 & + 0.03 & @xmath1330.00 & + 0.00 & & 0.07 & & 0.08 + cr i & + 0.10 & @xmath1330.04 & @xmath1330.01 & @xmath1330.00 & & 0.06 & & 0.12 + cr ii & @xmath1330.02 & + 0.05 & @xmath1330.00 & + 0.00 & & 0.07 & & 0.09 + mn i & + 0.08 & @xmath1330.07 & @xmath1330.03 & @xmath1330.01 & & 0.04 & & 0.12
+ fe i & + 0.08 & @xmath1330.04 & @xmath1330.01 & @xmath1330.01 & & 0.01 & & 0.09 + fe ii & + 0.01 & + 0.04 & @xmath1330.01 & + 0.00 & & 0.05 & & 0.07 + co i & + 0.08 & @xmath1330.08 & @xmath1330.03 & @xmath1330.01 & & 0.09 & & 0.15 + ni i & + 0.07 & @xmath1330.04 & @xmath1330.02 & @xmath1330.01 & & 0.06 & & 0.10 + zn i & + 0.04 & + 0.01 & + 0.00 & + 0.00 & & 0.07 & & 0.08 + sr ii & + 0.07 & @xmath1330.01 & @xmath1330.07 & @xmath1330.02 & & 0.07 & & 0.12 + y ii & + 0.05 & + 0.04 & @xmath1330.01 & + 0.00 & & 0.07 & & 0.10 + ba ii & + 0.08 & + 0.03 & @xmath1330.02 & @xmath1330.01 & & 0.06 & & 0.11 + + + na i & + 0.12 & @xmath1330.01 & @xmath1330.06 & @xmath1330.03 & & 0.07 & & 0.16 + mg i & + 0.06 & @xmath1330.06 & @xmath1330.02 & @xmath1330.01 & & 0.03 & & 0.09 + al i & + 0.05 & @xmath1330.08 & @xmath1330.06 & @xmath1330.04 & & 0.10 & & 0.16 + si i & + 0.09 & @xmath1330.11 & @xmath1330.03 & @xmath1330.02 & & 0.10 & & 0.18 + k i & + 0.08 & + 0.00 & @xmath1330.01 & + 0.00 & & 0.10 & & 0.13 + ca i & + 0.06 & @xmath1330.02 & @xmath1330.01 & @xmath1330.00 & & 0.03 & & 0.07 + sc ii & + 0.03 & + 0.06 & @xmath1330.02 & @xmath1330.00 & & 0.04 & & 0.08 + ti i & + 0.11 & @xmath1330.03 & @xmath1330.01 & @xmath1330.00 & & 0.02 & & 0.12 + ti ii & + 0.02 & + 0.06 & @xmath1330.02 & @xmath1330.00 & & 0.02 & & 0.07 + v i & + 0.11 & @xmath1330.04 & + 0.00 & + 0.00 & & 0.10 & & 0.15 + v ii & + 0.01 & + 0.05 & @xmath1330.01 & + 0.00 & & 0.05 & & 0.07 + cr i & + 0.10 & @xmath1330.03 & @xmath1330.01 & @xmath1330.01 & & 0.03 & & 0.11 + cr ii & @xmath1330.03 & + 0.07 & @xmath1330.01 & + 0.00 & & 0.07 & & 0.10 + mn i & + 0.09 & @xmath1330.05 & @xmath1330.03 & @xmath1330.02 & & 0.04 & & 0.12
+ fe i & + 0.08 & @xmath1330.03 & @xmath1330.02 & @xmath1330.01 & & 0.01 & & 0.09 + fe ii & @xmath1330.01 & + 0.07 & @xmath1330.01 & + 0.00 & & 0.03 & & 0.08 + co i & + 0.09 & @xmath1330.06 & @xmath1330.03 & @xmath1330.02 & & 0.10 & & 0.15 + ni
i & + 0.08 & @xmath1330.03 & @xmath1330.01 & @xmath1330.01 & & 0.03 & & 0.09 + zn i & + 0.04 & + 0.01 & @xmath1330.01 & + 0.00 & & 0.07 & & 0.08 + sr ii & + 0.04 & + 0.03 & @xmath1330.08 & @xmath1330.03 & & 0.07 & & 0.12 + y ii & + 0.05 & + 0.06 & + 0.00 & + 0.01 & & 0.07 & & 0.10 + ba ii & + 0.08 & + 0.07 & @xmath1330.05 & @xmath1330.02 & & 0.07 & & 0.13 | metal - poor stars in the milky way are local relics of the epoch of the first stars and the first galaxies
. however , a low metallicity does not prove that a star formed in this ancient era , as metal - poor stars form over a range of redshift in different environments .
theoretical models of milky way formation have shown that at constant metallicity , the oldest stars are those closest to the center of the galaxy on the most tightly - bound orbits .
for that reason , the most metal - poor stars in the bulge of the milky way provide excellent tracers of the chemistry of the high - redshift universe .
we report the dynamics and detailed chemical abundances of three stars in the bulge with @xmath0}\lesssim-2.7 $ ] , two of which are the most metal - poor stars in the bulge in the literature .
we find that with the exception of scandium , all three stars follow the abundance trends identified previously for metal - poor halo stars .
these three stars have the lowest [ /fe ] abundances yet seen in @xmath1-enhanced giant stars in the galaxy .
moreover , all three stars are outliers in the otherwise tight [ /fe][/fe ] relation observed among metal - poor halo stars .
theoretical models predict that there is a 30% chance that at least one of these stars formed at @xmath2 , while there is a 70% chance that at least one formed at @xmath3 .
these observations imply that by @xmath4 , the progenitor galaxies of the milky way had both reached @xmath0}\sim-3.0 $ ] and established the abundance pattern observed in extremely metal - poor stars . |
multiwavelenght follow - up studies of grbs in the last decade have established that a significant fraction of long grbs arise from the simultaneous collapse of a massive star ( e.g. woosley & bloom 2006 , della valle 2006 ) .
best examples of this are the sn / grb associations so far discovered in the local universe , 2006aj with grb 060218 ( pian 2006 , campana 2006 , modiaz et al .
2006 ) and sn 1998bw with 980425 ( galama 1998 ) .
further evidence ( @xmath12 ) comes from sn 2003dh with grb 030329 ( stanek 2003 ; hjorth 2003 ) and grb 031203 , which have been found to be associated with bright broad - lined type ic sne ( malesani 2004 ) .
this type of grb / sn association applies to a significant fraction of long grbs , but not all of them ( della valle et al .
2006a , fynbo et al .
2006 , gal - yam et al .
2006 , gehrels et al .
2006 ) . at larger redshifts
the association grb - sne relies on about a dozen of rebrightenings ( `` bumps '' ) observed during the late decay stages of the grb afterglow light curve ( see zeh 2004 and references therein ) .
they have been interpreted as due to the emergence of the optical contribution of an underlying sn ( bloom 1999 ) . in two cases , grb 021211/sn 2002lt ( della valle 2003 ) and grb 050525a / sn 2005nc ( della valle 2006b ) , spectroscopic observations obtained during these `` bumps '' are suggestive of the presence of sn components . in spite of these remarkable achievements ,
there are still many uncertainties hampering our knowledge the rate of these events .
particularly , the recent discovery of grb 060218 at z=0.03 has raised the question whether or not a population of `` local '' and `` low - luminosity '' grbs ( ll - grbs , i.e. @xmath0 erg / sec ) with different properties from the energetically `` high - luminosity '' grbs ( hl - grbs ) does exist ( e.g. cobb 2006 , soderberg et al . 2006a , amati et al .
2006 ) .
the aim of this letter is twofold : first we derive the rate for the local ( @xmath13 ) and ll - grbs , with two independent approaches . then , we investigate the existence of two classes of grbs by measuring the ratios of the rates of ll - grbs and hl - grbs to sne - ibc progenitors .
we consider three ll - grbs , grb 980425 ( z=0.008 ) , grb 060218 ( z=0.03 ) and grb 031203 ( z=0.105 ) detected by _ bepposax
_ wide field camera ( wfc ) , _ swift _ burst alert telescope ( bat ) and integral respectively and infer from them an empirical rate for subluminous grbs . following soderberg et al .
2006a we estimate the peak flux in the range 1 - 1000 kev and compare it with the threshold peak flux calculated by band ( 2003 , 2006 ) . in this way
we derive the maximum distance ( @xmath14 ) the event could be detected and therefore the maximum volume ( @xmath15 ) .
considering the sky coverage ( @xmath16 ) and the number of years of operation ( @xmath17 ) of each individual the detector the rate of ll - grbs similar to the grbs given above is : @xmath18 * grb 980425 * this burst was detected by _ bepposax _
wfc ( pian 1999 ) and by batse ( kippen 1998 ) . the peak flux in the 50 - 300 kev band was @xmath19 ph @xmath20 s@xmath3 .
the redshift of this burst is z=0.0085 , implying an isotropic peak luminosity @xmath21 erg s@xmath3 several orders of magnitude smaller than typical long hl - grbs that have @xmath22 erg@xmath23 on average . by using the spectrum given by jimenez , band & piran ( 2001 )
we extrapolate the peak flux @xmath24 ph @xmath20 s@xmath3 . given the threshold @xmath25 ph @xmath20 s@xmath3 we find @xmath26 mpc .
the bepposax sky coverage was about 0.08 and the operation time @xmath27 4 yrs .
therefore the rate of grb 980425-like events is @xmath28 gpc@xmath2 yr@xmath3 .
* grb 060218*. this burst was detected by _
the peak flux was @xmath29 erg @xmath20 s@xmath3 and the spectral index 1.5 ( campana 2006 ) .
the redshift is , z=0.03 , implying an isotropic peak luminosity of the order of @xmath30 .
the extrapolated peak flux is @xmath31 ph @xmath20 s@xmath3 .
given the threshold @xmath32 ph @xmath20 s@xmath3 we find @xmath33 mpc .
the _ swift _ sky coverage was 0.17 and the operation time is one year
. therefore we estimate that the rate of grb 060218-like events is @xmath34 gpc@xmath2 yr@xmath3 .
* grb 031203 * this burst was detected by integral ( malesani 2004 ) .
the peak flux in the 20 - 200 kev band was @xmath35 ph @xmath20 s@xmath3 and the spectral index is 0.8 ( @xmath36 ) .
the redshift of this burst is , z=0.105 , implying an isotropic peak luminosity @xmath37 erg s@xmath3 that is three order of magnitude smaller than canonical long bursts but two order higher than the previous ones .
the extrapolated flux is @xmath38 ph @xmath20 s@xmath3 .
given the threshold @xmath39 ph @xmath20 s@xmath3 ( mereghetti & gtz 2005 ) we find @xmath40 mpc .
the integral sky coverage is 0.5% and the operation time was 3 yrs
. therefore the rate of grb 031203-like events is @xmath41 gpc@xmath2 yr@xmath3 .
the redshift and the luminosity of this burst are intermediate between ll - grbs and hl - grbs and this suggest a continuity between the two classes . the occurrence of two events such as grb 980425 and grb 060218 , within @xmath42 mpc distance implies a rate of @xmath43 ll - grb gpc@xmath2 yr@xmath3 in agreement with soderberg et al .
2006a result .
the attached errors represent @xmath44 poissonian standard deviation ( gherels 1986 ) .
another potential source of uncertainty ( which is not included ) is the correction for beaming .
however , we point out that ll - grbs may be much less collimated than typical hl - grbs ( see next section ) . one way to estimate the jet opening angle , @xmath45 , is to look for `` achromatic - breaks '' in the afterglow light curve .
the steepening of the afterglow light curve , break , occurs when the bulk lorentz factor of the relativistic outflow becomes lower than @xmath46 . from the data on the afterglow of grb 980425 and grb 060218 it seems that the presence of breaks in the light curves is very late , if existent at all , implying very large @xmath45 .
for example radio observations of grb 060218 ( soderberg 2006a ) have set @xmath47 which corresponds to a beaming factor ( @xmath48 ) , @xmath49 . by interpreting the break in the x - ray light curve of grb 031203 as due to a jet
, we obtain @xmath50 , which corresponds to @xmath51 .
the beaming corrected rate would thus be @xmath52 event gpc@xmath2 yr@xmath3 , consistent with the value reported above .
the above estimate of the ll - grbs rate may be affected by the small number of available events , therefore an independent check is in order . to do this we compare the empirical rate with the rate inferred from the large grb samples obtained by both batse and swift .
as suggested by schmidt ( 2001 ) , the grb local rate can be estimated by using the batse peak flux distribution .
this is given by the convolution of two unknown quantities , the luminosity function ( lf ) and the grb formation rate . following guetta , piran & waxman ( 2005 ) and guetta et al .
( 2004 ) , we assume that grbs trace the star formation history and adopt the rowan - robinson star formation rate ( sfr ) , @xmath53 , ( rowan - robinson 1999 ) .
other rate evolutions have been analized in detail in firmiani et al .
then we find the best fit to the lf by comparing the batse observed peak flux distribution with the predicted one , assuming a form for the lf ( i.e. a power law with a minimum and maximum luminosities ) .
we show in fig .
1 that a single power - law lf , @xmath54 , with a minimum ( @xmath55 erg s@xmath3 ) and maximum luminosity ( l@xmath56 erg s@xmath3 ) fits the batse peak flux distribution very well . to obtain the observed local rate of grbs per unit volume , @xmath57 , we need to estimate the effective full - sky coverage of our grb sample .
the batse catalog represents 3.185 years of batse full sky coverage implying a rate of 692 grb per year . using our lf we find @xmath58gpc@xmath2yr@xmath3 , that is much smaller than what would be required in order to explain the local rate of ll - grbs . ] .
however the local rate discussed in the previous paragraph was derived under the hypothesis that classical bursts exceed by far the luminosity of grb 980425 , of grb 060218 and of grb 031203 ( i.e. in the estimate of the rate guetta , piran & waxman ( 2005 ) considered a minimum luminosity for the luminosity function ( lf ) @xmath59 erg / sec ) : as such , this rate can not be compared with the empirical one derived for ll - grbs .
a self consistent estimate can only be obtained with a lf that extends down to luminosities as low as that of grb 980425 and grb 060218 .
it is important to realize that most of the bursts below @xmath60 erg s@xmath3 ( ll - grbs ) are undetectable by current detectors , unless they are extremely close ( z@xmath61 ) .
on the basis of the observables that we have ( i.e. the peak flux distribution ) we can not constrain the minimum value of the lf , therefore we find reasonable to take the minimum luminosity observed in grbs ( @xmath30 ) as our minimum value for the lf .
if we repeat the same procedure given above using the same lf and l@xmath62 but a different @xmath63 , we find that the batse peak flux distribution is fitted very well and the rate increases as @xmath64gpc@xmath2yr@xmath3 .
the increase in the rate of ll - grb is due to the fact that the luminosity function increases rapidly with decreasing luminosity from @xmath65 ergs / sec down to @xmath63 ( guetta & piran 2006 ) . if we further extend the lf considering @xmath66 in the range between @xmath67 we obtain @xmath68 gpc@xmath2 yr@xmath3 consistent with the empirical rate determined in section 2 .
this luminosity function fits well also the swift peak flux distribution . taking into account that swift has detected 122 long grbs in 1.6 yr and covers 1/6 of the sky ,
applying the same method described above , we find a rate of @xmath69 for @xmath66 in the range between @xmath67 consistent with the rate derived from the batse data and with the empirical rate .
the fraction of sne - ibc that produces grbs can be measured as follows .
a rate of @xmath70 sne - ibc gpc@xmath2 yr@xmath3 is derived by combining the local density of b luminosity of @xmath71 mpc@xmath2 ( e.g. madau , della valle & panagia 1998 ) with the rate of sne - ibc observed in sbc
irr hubble types ( these morphological types are appropriate to represent the grb hosts ) of 0.16 sne per century and per 10@xmath72 @xmath73 ( snu units , cappellaro , evans & turatto 1999 ) .
this sn rate has to be compared with the rate of hl - grbs of @xmath74 grb gpc@xmath2 yr@xmath3 ( guetta , piran & waxman 2005 ) after rescaling for the jet beaming factor , @xmath75 .
there exist different estimates for this parameter : from @xmath76 ( guetta , piran & waxman 2005 ) to @xmath77 ( frail 2001 ) corresponding to beaming angles @xmath78@xmath79 , respectively . taking these figures at their face value
, we find the ratio hl - grb / sne - ibc to be in the range : @xmath80 and the ratio ll - grbs / sne - ibc : @xmath5 ( for @xmath81 ) .
radio surveys give independent and consistent constraints : berger ( 2003 ) find that the incidence of sn 1998bw - like events , in the nearby universe , is @xmath82 , soderberg et al .
( 2006c ) find hl - grb / sne - ibc @xmath83 .
the computation of the grb / hne ratio requires a further step .
the measurement of the sn rate is based on the control - time methodology ( zwicky 1938 ) that implies the systematic monitoring of galaxies of known distances and the use of appropriate templates for the light curves of each sn type ( see cappellaro 1993 for bias and uncertainties connected with this procedure ) .
unfortunately all hypernovae reported in tab .
i have not been discovered during time ` controlled ' surveys , and therefore any attempt to derive an absolute value of the rate of hypernovae should be taken with caution .
one possibility is to compute the frequency of occurrence of all sne - ib / c and hypernovae in a limited distance sample of objects .
we extracted 193 sne - ib / c from an upgraded version of the asiago catalog ( ` http://web.pd.astro.it/supern ` ) , 19 of these have been spectroscopically classified as hypernovae . in fig .
2 we show that the cumulative distributions of the recession velocities of the hosts of hne ( i.e. broad - lined sne - ib / c ) and `` standard '' sne - ibc ( dotted line ) are statistically indistinguishable ( ks probability=0.73 ) for @xmath84 km / s . within this volume
we find 158 objects , 12 of which are hne . after excluding sn 1998bw and 2006aj , because they were not serendipitously discovered and assuming that the host galaxies of both ` normal ' sne ib / c and hypernovae were monitored with a comparable level of efficiency , we infer that the fraction of hypernovae is about @xmath85 of the total number of sne ib / c .
therefore the hn rate turns out to be @xmath27 0.015 sne per century and per 10@xmath72 @xmath73 .
note that this value is about an order of magnitude larger than reported by podsiadlowski et al .
these authors find a similar ratio hne / sne - ibc ( about 5% ) , however they assume that hne are on average 4 times brighter than `` standard '' sne - ibc , and therefore they can be detected in a volume about 10 times larger , therefore they adopt an `` actual '' ratio hne / sne - ibc @xmath86 .
however their assumption on the luminosity at maximum of hne does not seem supported by either our fig . 2 and the results of soderberg et al .
( 2006b ) who find the magnitude distribution of grb / xrf - sne and local sne - ib / c ( see richardson et al .
2006 ) to be statistically indistinguishable .
finally we check whether our result is consistent with the lack of a single hn detection in the cappellaro s ( 1999 ) sample . in particular , from poisson statistics we find that the probabilities of obtaining a null result are @xmath2750% and @xmath275% when the expected hn numbers are 0.7 and 3.0 . after
using the control time of cappellaro ( 1999 ) we derive @xmath87 and @xmath88 hne per century per 10@xmath72 @xmath73 .
then , one should expect that the hn rate is likely similar to the 50% probability value ( 0.012 snu ) and hardly higher than the 5% probability value ( 0.05 snu ) .
the rate of 0.015 snu derived in this paper is fully consistent with the limits set by the poissonian statistics .
together these data imply a ratio ll - grbs / hne in the range @xmath89 and beaming factors @xmath90 ( or @xmath91 ) . in view of these findings , we shall assume , in the following , that ll - grbs emit ( almost ) isotropically .
it is interesting to compare the beaming corrected rate of hl - grbs with the rate of ll - grbs . for every observed hl - grb
there are @xmath92 bursts that are not observed , thus the true rate is @xmath93 gpc@xmath2 yr@xmath3 . for isotropic ll - grbs
our analysis yields comparable rates : @xmath94 gpc@xmath2 yr@xmath3 ( see section 2 ) , @xmath95 gpc@xmath2 yr@xmath3 ( from swift data , section 3 ) and @xmath96 gpc@xmath2 yr@xmath3 ( from batse data , section 3 ) .
these results indicate that if there are two populations of grbs ( hl - grbs and ll - grbs ) the respective frequency of occurrence are comparable within a factor of @xmath97 .
however the data do not exclude the case that only one population of grbs is responsible for giving rise to both the highly collimated component ( preferentially observable at high z ) and to the almost isotropic components ( detectable only in nearby grbs ) ( e.g. woosley & heger 2006 ) . .hypernovae
[ cols="^ , > , > " , ]
from the available information on grbs and sne - ibc rates a number of interesting results emerge : \i ) we have computed the hn rate , on robust empirical grounds , and found that this sub - class of sne - ibc , characterized by broad - lined spectra , includes about 7% of sne - ibc .
an analysis of the cumulative distribution of the recession velocities of the respective hosts , does not suggest that hne are intrinsically more luminous ( on average ) than standard sne - ibc .
all together these facts imply a hn rate of @xmath7 hne @xmath98 .
\ii ) the ratio hl - grbs / hne is smaller than 1 , possibly in the range 0.040.3 .
the ratio ll - grb / hne is in the range 0.11 .
\iii ) if we assume that all hne are able to produce ll - grbs , then the ll - grb / hne ratio allows us to constrain the beaming factor to be smaller than @xmath90 .
\iv ) from the analysis of two nearest grbs we have derived a ll - grb rate of @xmath43 grbs gpc@xmath2 yr@xmath3 .
the attached errors represent @xmath44 poissonian standard deviation .
this result is consistent with the rates of 2001800 and 1001200 ll - grb gpc@xmath2 yr@xmath3 , derived from the batse and swift grbs .
\v ) the frequencies of occurrence of hl - grbs ( hl - grb / sne - ibc @xmath82 ) and ll - grbs ( ll - grb / sne - ibc @xmath99 ) may suggest the existence of two physically distinct classes of grbs ( e.g. cobb et al .
2006 , soderberg et al .
2006a , pian et al .
2006 , amati et al .
2006 ) in which ll - grbs are ( intrinsically ) more frequent events than hl - grbs .
however , due to the uncertainties , we can not exclude that a single population of grbs originates both the isotropic and sub - energetic component , detectable only in nearby grbs , and the highly collimated component , observable by sampling huge volumes of space , and thus mainly detectable in high - z grbs .
we thank the anonymous referee for useful comments which have improved the presentation of this paper and sandro mereghetti , daniele malesani and lorenzo amati for useful discussions .
the authors are also indebted with luigi stella and mario vietri for their critical reading of the manuscript .
m.d.v . is grateful to the university of tokyo for the friendly hospitality and creative atmosphere .
mdv acknowledges the prin - inaf 2005 grant for financial support .
amati et al .
2006 , a&a , in press band , d. , 2003 , apj 588 , 945 .
band , d. , 2006 , apj 644 , 378 .
berger , e. , kulkarni , s. r. , frail , d. a. , soderberg , a. m. 2003 , apj , 599 , 408 bloom , j.s . , kulkarni , s.r . , djorgovski , s.g . 1999 ,
nature , 401 , 453 campana , s. , 2006 , nature , 442 , 1008 cappellaro , e. , turatto , m. , benetti , s. , tsvetkov , d. y. , bartunov , o. s. , makarova , i. n. 1993 , a&a , 268 , 47 cappellaro , e. , evans , r. , & turatto , m. 1999 , a&a , 351 , 459 cobb , b.e . , bailyn ,
c.d . , van dokkum , p.g . , & natarajan , p. 2006 ,
apjl , 645 , l113 della valle , m. 2003 , a&a 406 , l33 .
della valle , m. 2006 , in gamma - ray bursts in the swift era , sixteenth maryland astrophysics conference , held in washington , november , 2005 , eds .
holt , n. gehrels , and j.a .
aip conference proceedings , vol .
836 , p.367 - 379 della valle , m. 2006a , nature , in press , astro - ph/0608322 della valle , m. , malesani , d. , bloom , j.s . , 2006b , apj , 642 , l103 filippenko , a.v .
1997 , iauc , n. 6783 filippenko , a.v . , leonard , d.c . , moran , e.c . 2002 , iauc , n.7845 filippenko , a.v .
& chornock , r. 2003a , iauc , n. 8084 filippenko , a.v . ,
foley , r.t . , swift , b. 2003b , iauc , n. 8234 filippenko , a. v. , foley , r. j. , matheson , t. 2005 , iauc , n. 8639 firmani , c. et al .
2004 , apj 611 , 1033 foley , r.j . ,
papenkova , m.s . , swift , b. j. 2003 , pasp , 115 , 1220 foley , r.j . ,
wong , d.s . , moore , m. & filippenko , a.v .
2004 , iauc , n. 8353 frail , d.a . ,
2001 , apj , 562 , l55 frieman , j. 2005a , iauc n.8616 frieman , j. 2005b , iauc n.8634 fynbo , j. et al .
2006 , nature , in press , astro - ph/0608313 galama , t.j . ,
vreeswijk , p.m. , van paradijs , j. , 1998 , nature , 395 , 67 gal - yam , a. et al .
2006 , nature , in press , astro - ph/0608257 gehrels , n. 1986 , apj , 303 , 336 gehrels , n. et al .
2006 , nature , in press , astro - ph/0610635 guetta , d. , perna , r. , stella , l. , & vietri , m. 2004 , apj , 615 , l7 guetta , d. , piran , t. , & waxman , e. 2005 , apj , 619 , 412 guetta , d. & piran , t. 2006 , a&a 453 , 823 hatano , k. , branch , d. , nomoto , k. , deng j.s . , maeda , k. , nugent p. , aldering , g. 2001 , 198th baas , 33 , p. 838
hjorth 2003 , nature 423 , 847 .
jimenez , r. , band , d. & piran , t. , 2001 , apj 561 , 171 kippen , m. , 1998 , gcn 67 .
madau , p. , della valle , m. , & panagia , n. 1998 , mnras , 297 , l17 malesani , d. tagliaferri , g. , chincarini , g. , 2004 , apj , 609 l5 .
masetti , n. , 2006 , gcn 4803 matheson , t. , challis , p. & kirshner , r. 2003 , iauc , n. 8234 mazzali , p. , deng , j. , maeda , k. , 2002 , apj , 572 , l61 mazzali , p. a. , deng , j. , maeda , k. , nomoto , k. , filippenko , a. v. , matheson , t. 2004 , apj , 614 , 85 mereghetti , s. , & gtz , d. 2005 , nuovo cimento c , 28 , 259 ; http://ibas.iasf-milano.inaf.it/ modjaz , m. , challis , p. , kirshner , r. , matheson , t. , berlind , p. 2005 , iauc n. 8575 modjaz , m. 2006 , apjl , 645 , l21 pian , e. 1999 , a&a 138 , 407 pian , e. 2006 , nature , 442 , 1011 podsiadlowski , p. , mazzali , p. , nomoto , k. , lazzati , d. , & cappellaro , e. 2004 , apj , 607 , l17 porciani , c. & madau , p. 2001 , apj 548 , 522 .
richardson , d. , branch , d. , baron , e. 2006 , 131 , 2233 riello , m. 2004 , iauc 8296 roman , b. , & rostopchin , s. 2006 , iauc , n.8657 rowan - robinson , m. 1999 , ap&ss , 266 , 291 schmidt , m. 2001 , apj 552 , 365 .
soderberg , a. m. 2006a , nature , 442 , 1014 soderberg , a.m. 2006b , apj , 636 , 391 soderberg , a.m. , nakar , e. , kulkarni , s.r .
, & berger , e. 2006c , apj , 638 , 930 stanek 2003 , apj 591 , l17 .
panagia , n. , rupen , m. p. , paczynski , b. 2004 , iauc 8282 woosley , s.e .
, & bloom , j. 2006 , ara&a , 44 , 507 woosley , s.e . , & heger , a. 2006 , apj , 637 , 914 zeh , a. , klose , s. , & hartmann , d.h .
2004 , apj , 609 , 952 zwicky , f. 1938 , apj , 88 , 529 | we measure the local rates of `` low - luminosity '' ( ll - grbs , i.e. @xmath0erg / sec ) and `` high - luminosity '' gamma - ray bursts ( hl - grbs ) .
the values are in the range @xmath1 gpc@xmath2 yr@xmath3 and @xmath4 gpc@xmath2 yr@xmath3 , respectively , and the ratios to sne - ibc @xmath5 and @xmath6 . these data may suggest the existence of two physically distinct classes of grbs in which ll - grbs are ( intrinsically ) more frequent events than hl - grbs . however , with the present data we can not exclude the possibility of a single population of grbs which give rise to both an isotropic low - luminous emission ( ll - grbs : detectable only in nearby grbs ) and to a highly collimated high - luminous emission ( hl - grbs : detectable preferentially at high - z ) .
we compute also the rate of sne - ibc characterized by broad - lined spectra ( hypernovae ) and found it to be about @xmath7hne yr@xmath3 @xmath8 l@xmath9 ( i.e less than @xmath10 of sne - ibc occurring in spirals ) .
this result implies that the ratio hl - grbs / hne is smaller than 1 , possibly in the range 0.040.3 .
we have used the ratio between hypernovae and ll - grbs to constrain their beaming factor to @xmath11 or less . 0 _ 5_,0.5 |
the study of extra - solar planets has been driven forward by the discovery and detailed analysis of those planets that transit the disc of their host star .
these transiting systems provide information on the planet s radius ( e.g. * ? ? ?
* ) and absolute mass , information that is vital for the accurate characterisation of both the planet and star .
planets transiting bright stars ( @xmath18 ) are particularly prized ; they enable detailed follow - up studies that can provide confirmation of the planets existence , as well as allowing the exploration of the planets characteristics , such as spin - orbit alignment and atmospheric composition .
the space - based kepler @xcite transit survey mission has vastly expanded the size of the exoplanet population , and has discovered small , earth - like planets , circumbinary planets , and multi - planet systems that are inaccessible to instruments on the ground .
it is the ground - based surveys though , such as the wide angle search for planets ( wasp ; @xcite ) , hatnet @xcite , tres @xcite , xo project @xcite , and kelt @xcite , that are leading the way in the discovery of planets orbiting bright stars .
wasp is by far the most successful of these surveys , with more than 100 published planet discoveries to date .
a recent study of f , g , and k stars in the sloan digital sky survey ( sdss ) showed that approximately @xmath19percent of solar - type stars have binary companions with periods of @xmath20days @xcite .
this supports earlier results by @xcite ( @xmath21percent solar type stars in binaries ) , @xcite , and @xcite , among others .
however , the binary fraction seems to be lower for exoplanet systems @xcite , indicating a suppression of planet formation @xcite .
there is a strong selection effect acting against binary systems in planet search programs , that @xcite acknowledge and @xcite correct for .
the presence of a companion star to the planet host introduces the problem of light from said star contaminating either the photometric observations ( diluting the transit depth ) , spectroscopic observations ( introducing a second set of spectral lines , and thus a second cross - correlation function peak ) , or both .
the level of contamination depends on several factors - the resolution of the instrument , aperture size , fibre diameter , seeing , the separation of the stars , and the magnitude difference between the stellar components .
these factors vary considerably from system to system , and the presence of a binary companion to an exoplanet candidate star tends to reduce the likelihood of that candidate becoming the target of follow - up observations .
nevertheless , there are several examples of hot jupiter exoplanets in s - type orbits around binary stars , including wasp-70ab @xcite ; wasp-77ab @xcite ; wasp-94ab and bb @xcite ; kelt-2ab @xcite , and kepler-14ab @xcite .
analysis of the known planets in binary systems deal with the contamination in different ways . for their analysis of wasp-70ab
, @xcite corrected the photometric data using flux ratios derived from their observations .
a similar approach was taken by @xcite for wasp-94ab , and by @xcite for wasp-77a .
kepler-14ab is possibly the most similar system to that we consider here , and @xcite account for the dilution by placing a prior on the flux ratio of the binary components .
the most complex approach is that of @xcite for kelt-2ab , who carry out an iterative combination of markov chain monte carlo analysis and sed fitting to constrain the transit depth . in this paper
we present the discovery of wasp-85ab , a transiting hot jupiter exoplanet with mass @xmath22@xmath2 and radius @xmath23@xmath4 .
the solar - type host star , wasp-85a is the brighter component of a close visual binary , bd+07@xmath242474 , while the companion star , wasp-85b , is cooler than the host star , but of similar magnitude .
the presence of the companion star complicates the standard analysis procedure used for wasp planets ; we account for the 3rd light contamination by adding an additional term to our stellar model , derived from our photometric data and controlled by a markov chain monte carlo jump parameter and prior.there is a clear , long - term trend in the position angle between the binary components , which indicate an orbit of @xmath25years for the binary .
the system lies in the field of view of k2 campaign1 , and was observed in both long and short cadence modes .
the k2 light curve clearly shows the planetary transits , along with the signature of stellar variability .
analysis of the k2 data from both observing modes is ongoing , and will be presented in a follow - up paper . in section[sec :
obs ] we detail the photometric and spectroscopic observations of the wasp-85 system , while in section[sec : analysis ] we discuss the methods by which we derive the system parameters .
we discuss the characteristics of the star and planet in section[sec : discuss ] , and conclude by summarising our results in section[sec : summary ] . throughout the manuscript we use the tdb time standard and barycentric julian dates in our analysis , following the advice set out in @xcite .
our results are based on the equatorial solar and jovian radii , and masses taken from allen s astrophysical quantities .
for a detailed account of the wasp telescopes , observing strategy , data reduction , and candidate identification and selection procedures , see @xcite and @xcite .
bd+07@xmath242474 lies near the celestial equator , and was thus observed by both superwasp - north ( located at the observatorio del roque de los muchachos on la palma , spain ) and wasp - south ( located at the south african astronomical observatory near sutherland , south africa ) .
these observations resulted in @xmath26 data , spanning the period 2008 - 02 - 05 to 2011 - 03 - 29 .
the system was first identified as a planet candidate from superwasp - north data in 2008 .
initial analysis revealed transit like features with an apparent period of @xmath27days .
this signal was subsequently identified in both the superwasp - north and wasp - south photometry independently , and the star was selected for photometric and spectroscopic follow - up observations .
subsequent to these observations , the secondary peak in the periodogram , at @xmath28days , was found to be the correct period .
the radius of the synthetic aperture ( @xmath29 ) used to extract the flux of bd+07@xmath242474 is much greater than the maximum binary separation ( @xmath30 ; see section[sec : binary ] ) , such that the wasp lightcurve includes flux contributions from both components of the binary .
initial spectroscopic reconnaissance was carried out between 2008 and 2010 using the sophie spectrograph @xcite mounted on the @xmath31-m telescope of the observatoire de haute - provence ( ohp ) , resulting in the acquisition of @xmath32 spectra .
rvs were derived through cross - correlation with a spectral mask suitable for a star of g2 spectral type .
the separation of the binary components is smaller than the fibre diameter of sophie ( @xmath33 ) , and thus the radial velocities ( rvs ) obtained from this instrument are contaminated by the contribution of the companion .
these data show a sinusoidal variation in radial velocity ( rv ) with a period of @xmath28days , in disagreement with the period identified in the wasp lightcurves .
further spectroscopic observations were made using the fibre - fed coralie spectrograph at the euler - swiss telescope at eso s la silla observatory , and using harps at the eso @xmath34 m telescope , also at la silla .
a total of @xmath35 observations were made using coralie between 2009 january 03 and 2014 june 24 . for details of the instrument and data reduction procedure ,
see @xcite and @xcite .
rvs were derived using cross - correlation with a spectral mask suitable for a g2-type star , and confirmed the @xmath28day period as the correct one . unfortunately , while the coralie guiding camera is able to identify the presence of both stellar components , the aperture of the spectrograph ( @xmath36 ) is insufficiently small to access a single stellar component at a time .
the coralie spectra are therefore also contaminated by light from the companion star .
the level of contamination is seeing dependent owing to the close similarity of the aperture size and binary separation , rendering a correction to the rvs both uncertain and difficult to make .
eight observations of the brighter binary component ( hereafter wasp-85a ) , and five observations of the fainter companion star ( hereafter wasp-85b ) were made using harps .
the small spectrograph aperture ( @xmath37diameter ) and good seeing allowed the components to be observed separately , with little contamination from the other star .
rvs were extracted in the same manner as for the coralie observations .
we check for false positives induced by background eclipsing binaries or stellar activity by examining the bisector span measurements for our data .
we find no significant correlation with the radial velocities or the radial velocity residuals , although the observations for wasp-85b do seem to show a negative correlation between the parameters .
this is indicative of stellar activity @xcite .
we also plot the full widths at half maximum ( fwhm ) of our radial velocity observations as a function of orbital phase , checking for variations in phase @xcite . none
are seen ( see figure[fig : bisector ] ) .
we confirm which binary component is the planet host by checking for phasing of the harps radial velocity observations .
the observations of wasp-85a show variations in phase with both the photometry , and with the coralie and sophie observations , while the observations of wasp-85b show variations that are not correlated with the phase determined from the lightcurve . .
harps data for wasp-85a are denoted by solid , black triangles , data for wasp-85b by open , black triangles .
coralie data are denoted by solid , blue squares .
sophie data are denoted by solid , red circles .
only the wasp-85b data show a negative correlation between the two parameters . _
lower panel : _ radial velocity full width at half maximum ( fwhm ) as a function of orbital phase .
the uncertainties in the fwhms are taken to be @xmath38 .
the fwhm values have been offset from the minimum value for each data set to allow comparison .
there is no clear variation with orbital phase .
+ + , scaledwidth=48.0% ] we list our rv data in table[tab : rvs ] , along with the bisector spans , which measure the asymmetry of the cross - correlation functions ( ccfs ) , and the fwhms of the ccfs .
uncertainties on the bisectors were taken to be twice the uncertainty on the rv , while those on the fwhms were taken to be @xmath39 times the rv uncertainty .
lllllll + @xmath40 & rv & @xmath41 & bisector & @xmath42 & fwhm & @xmath43 + @xmath44 & & & & & & + ( days ) & ( ) & ( ) & ( ) & ( ) & ( ) & ( ) + + + @xmath45 & @xmath46 & @xmath47 & @xmath48 & @xmath49 & @xmath50 & @xmath51 + @xmath52 & @xmath53 & @xmath54 & @xmath55 & @xmath56 & @xmath57 & @xmath58 + @xmath59 & @xmath60 & @xmath61 & @xmath62 & @xmath63 & @xmath64 & @xmath65 + @xmath66 & @xmath67 & @xmath68 & @xmath69 & @xmath70 & @xmath71 & @xmath72 + @xmath73 & @xmath74 & @xmath75 & @xmath61 & @xmath76 & @xmath77 & @xmath78 + @xmath79 & @xmath80 & @xmath75 & @xmath81 & @xmath76 & @xmath82 & @xmath78 + @xmath83 & @xmath84 & @xmath61 & @xmath85 & @xmath63 & @xmath86 & @xmath65 + @xmath87 & @xmath88 & @xmath68 & @xmath89 & @xmath70 & @xmath90 & @xmath72 + @xmath91 & @xmath92 & @xmath61 & @xmath93 & @xmath63 & @xmath94 & @xmath65 + @xmath95 & @xmath96 & @xmath97 & @xmath85 & @xmath98 & @xmath99 & @xmath100 + @xmath101 & @xmath102 & @xmath97 & @xmath103 & @xmath98 & @xmath104 & @xmath100 + + + @xmath105 & @xmath106 & @xmath107 & @xmath108 & @xmath109 & @xmath110 & @xmath111 + @xmath112 & @xmath113 & @xmath114 & @xmath115 & @xmath116 & @xmath117 & @xmath118 + @xmath119 & @xmath120 & @xmath121 & @xmath122 & @xmath123 & @xmath124 & @xmath125 + @xmath126 & @xmath127 & @xmath128 & @xmath129 & @xmath130 & @xmath131 & @xmath132 + @xmath133 & @xmath134 & @xmath135 & @xmath136 & @xmath137 & @xmath138 & @xmath139 + @xmath140 & @xmath141 & @xmath121 & @xmath142 & @xmath123 & @xmath143 & @xmath125 + @xmath144 & @xmath145 & @xmath146 & @xmath147 & @xmath148 & @xmath149 & @xmath150 + @xmath151 & @xmath152 & @xmath153 & @xmath154 & @xmath155 & @xmath156 & @xmath157 + @xmath158 & @xmath159 & @xmath160 & @xmath161 & @xmath162 & @xmath163 & @xmath164 + @xmath165 & @xmath166 & @xmath167 & @xmath168 & @xmath169 & @xmath170 & @xmath171 + @xmath172 & @xmath173 & @xmath174 & @xmath175 & @xmath176 & @xmath177 & @xmath178 + @xmath179 & @xmath180 & @xmath181 & @xmath182 & @xmath183 & @xmath184 & @xmath185 + @xmath186 & @xmath187 & @xmath188 & @xmath189 & @xmath190 & @xmath191 & @xmath192 + @xmath193 & @xmath194 & @xmath195 & @xmath196 & @xmath197 & @xmath198 & @xmath199 + @xmath200 & @xmath201 & @xmath202 & @xmath203 & @xmath204 & @xmath205 & @xmath206 + @xmath207 & @xmath208 & @xmath209 & @xmath210 & @xmath211 & @xmath212 & @xmath213 + @xmath214 & @xmath215 & @xmath216 & @xmath217 & @xmath218 & @xmath219 & @xmath220 + @xmath221 & @xmath222 & @xmath223 & @xmath224 & @xmath225 & @xmath226 & @xmath227 + @xmath228 & @xmath229 & @xmath230 & @xmath231 & @xmath232 & @xmath233 & @xmath234 + @xmath235 & @xmath236 & @xmath237 & @xmath238 & @xmath239 & @xmath240 & @xmath241 + @xmath242 & @xmath243 & @xmath244 & @xmath245 & @xmath246 & @xmath247 & @xmath248 + @xmath249 & @xmath250 & @xmath251 & @xmath252 & @xmath253 & @xmath254 & @xmath255 + @xmath256 & @xmath257 & @xmath258 & @xmath259 & @xmath260 & @xmath261 & @xmath262 + @xmath263 & @xmath264 & @xmath265 & @xmath266 & @xmath267 & @xmath268 & @xmath269 + @xmath270 & @xmath271 & @xmath272 & @xmath273 & @xmath274 & @xmath275 & @xmath276 + @xmath277 & @xmath278 & @xmath279 & @xmath280 & @xmath281 & @xmath282 & @xmath283 + @xmath284 & @xmath285 & @xmath286 & @xmath287 & @xmath220 & @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 + @xmath332 & @xmath333 & @xmath334 & @xmath335 & @xmath336 & @xmath337 & @xmath338 + @xmath339 & @xmath340 & @xmath341 & @xmath342 & @xmath343 & @xmath344 & @xmath345 + @xmath346 & @xmath347 & @xmath348 & @xmath349 & @xmath350 & @xmath351 & @xmath352 + @xmath353 & @xmath354 & @xmath355 & @xmath356 & @xmath357 & @xmath358 & @xmath359 + @xmath360 & @xmath361 & @xmath362 & @xmath363 & @xmath364 & @xmath365 & @xmath366 + @xmath367 & @xmath368 & @xmath369 & @xmath370 & @xmath371 & @xmath372 & @xmath373 + + + @xmath374 & @xmath375 & @xmath376 & @xmath377 & @xmath378 & @xmath379 & @xmath380 + @xmath381 & @xmath382 & @xmath383 & @xmath384 & @xmath385 & @xmath386 & @xmath387 + @xmath388 & @xmath389 & @xmath167 & @xmath390 & @xmath169 & @xmath391 & @xmath171 + @xmath392 & @xmath393 & @xmath394 & @xmath395 & @xmath396 & @xmath397 & @xmath398 + @xmath399 & @xmath400 & @xmath401 & @xmath402 & @xmath403 & @xmath404 & @xmath405 + follow - up photometric observations to solidify the planet s ephemeris , and check for transit timing shifts , were carried out using the james gregory telescope ( jgt ) at the university of st andrews observatory , the robotic transiting planets and planetesimals small telescope ( trappist ; @xcite ) at la silla , the euler - swiss telescope using eulercam @xcite , also at la silla , and the liverpool telescope ( lt ; @xcite ) on la palma using the io : o instrument .
these observations are summarised in table[tab : phot ] . lllll + instrument & date & time of observations & filter & @xmath406 + + jgt & 2009 - 03 - 20 & @xmath407 & cousins r & @xmath408 + io : o & 2013 - 01 - 12 & @xmath409 & sloan @xmath410 & @xmath411 + io : o & 2013 - 01 - 28 & @xmath412 & sloan @xmath410 & @xmath411 + eulercam & 2013 - 03 - 23 & @xmath413 & rg & @xmath414 + trappist & 2013 - 03 - 31 & @xmath415 & sloan @xmath410 & @xmath416 + trappist & 2013 - 04 - 16 & @xmath417 & sloan @xmath410 & @xmath418 + trappist & 2013 - 05 - 09 & @xmath419 & sloan @xmath410 & @xmath420 + trappist & 2013 - 05 - 25 & @xmath421 & sloan @xmath410 & @xmath422 + + * note : * the rg filter used in eulercam is a modified broad gunn - r filter , with a central wavelength of @xmath423 nm .
+ a slewing problem affected trappist during the transit on 2013 - 03 - 31 , leading to gaps in coverage .
the resolutions and pixel scales of the telescopes and instruments used for photometric follow - up were insufficient to distinguish the companion star .
all lightcurves are therefore of the blended combination of light from the two stellar components .
we co - added all of the individual harps spectra of wasp-85a and wasp-85b to produce a single spectrum with a typical s / n of around 60:1 and 30:1 , respectively .
the standard pipeline reduction products were used in the analysis , which was performed using the methods given in @xcite .
the effective temperature ( ) was determined from the excitation balance of the fe i lines .
the na i d lines , the cai line at 6439 , and the ionisation balance of fe i and fe ii were used as surface gravity ( ) diagnostics . the metallicity was determined from equivalent width measurements of several unblended lines . a value for microturbulence ( ) was determined from fe i using the method of @xcite .
the projected stellar rotation velocity ( ) was determined by fitting the profiles of several unblended lines .
values for macroturbulence ( ) of @xmath424 and @xmath425 were determined from the calibration of @xcite .
a best fit value of @xmath426 and @xmath427 were obtained for wasp-85a and wasp-85b , respectively .
the results are summarised in table[tab : star_params ] ; the quoted uncertainties account for the errors in and as well as the additional scatter induced by measurement and atomic data uncertainties .
the spectral type was estimated from using tableb.1 in gray ( 2008 ) , while the stellar mass and radius estimates were derived using the @xcite calibration .
llll + parameter & wasp-85a & wasp-85b & units + + ra & & j2000 + dec & & j2000 + v & @xmath428 & @xmath429 & tycho + & @xmath430 & @xmath431 & + distance & & pc + spectral type & g5 & k0 & + & @xmath432 & @xmath433 & k + mass , @xmath434 & @xmath435 & @xmath436 & @xmath10 + radius , @xmath437 & @xmath438 & @xmath439 & @xmath440 + & @xmath441 & @xmath442 & cgs + & @xmath443 & @xmath444 & + & @xmath424 & @xmath425 & + & @xmath445 & @xmath427 & + @xmath446 & & days + @xmath447 & @xmath448 & @xmath449 & + @xmath8}$ ] & @xmath450 & @xmath451 & + @xmath452 & @xmath453 & @xmath454 & + + * note : * masses and radii estimated using the calibration of @xcite .
spectral type estimated from using tableb.1 in @xcite .
abundances are relative to the solar values obtained by @xcite .
the rotation period is the mean of the periods determined from a periodogram analysis of the wasp lightcurve . using the harps spectra with @xmath455 for the core of the caiih+k lines , we calculate the @xmath452 activity index using the emission in the cores of the caiih+k lines following @xcite .
we find an activity index of @xmath456 for wasp-85a . for wasp-85b
we find @xmath457 using spectra with @xmath458 .
we carried out a search for stellar modulation in the wasp lightcurve using the method described in section4 of @xcite .
the wasp data were separated according to the year of observation , as starspot induced variability is not expected to be coherent on long timescales owing to the short lifetime of magnetic features on the the stellar surface .
the transit features were subtracted using the best fit model transit model for that year .
a modified , generalised lomb - scargle periodogram ( e.g. * ? ? ? * ) , computed using the method of @xcite over 4096 uniformly spaced frequencies from @xmath459 to @xmath460cycles / day , was used to search for significant periodicity in the lightcurves .
the resulting periodograms are shown in figure[fig : pgram ] , and the results for the most significant peak are given in table[tab : pgram ] . false alarm probabilities ( fap ) were calculated using a monte carlo bootstrap method @xcite . in the 2009 , 2010 , and 2011 data there are significant peaks at approximately @xmath461days .
we associate this with the stellar rotation of one of the stellar components .
the shorter period found in the 2008 data is approximately @xmath462 , and easily explained by the presence of two active regions on opposite sides of the star .
this would produce a photometric signature at twice the rotational frequency , and will contribute power in other seasons as well , but less dominantly .
note that there is significant power at periods close to @xmath37day in all four seasons .
this likely arises as a result of the diurnal observing schedule forced upon wasp by its ground - based nature .
we estimate a mean rotation period of @xmath463days using the results from all four seasons .
this value is in good agreement with the rotation period of @xmath464d implied by the values of and the stellar radius of wasp-85a .
however it is also in agreement with the rotation period of @xmath465days implied for wasp-85b , making it impossible to determine which star the rotational modulation originates from using this data alone .
we will discuss this result further in section[sec : discuss ] .
we whiten the four sets of data by fitting a harmonic series with @xmath466 as the fundamental period .
the number of terms in the series was determined by minimising the bayesian information criterion @xcite .
the resulting fit was divided - out , and the lightcurves concatenated to produce a whitened wasp lightcurve .
lllll + year & @xmath406 & period & amp & fap + & & ( days ) & & + + @xmath467 & @xmath468 & @xmath469 & @xmath470 & @xmath471 + @xmath472 & @xmath473 & @xmath474 & @xmath470 & @xmath475 + @xmath476 & @xmath477 & @xmath478 & @xmath470 & @xmath479 + @xmath480 & @xmath481 & @xmath482 & @xmath470 & @xmath483 + + adopted & & @xmath463 & & + + lithium is detected in wasp-85a , with an equivalent width of @xmath484m , corresponding to an abundance of @xmath485 .
this implies an age of between @xmath486 and @xmath487gyr .
there is no significant detection of lithium in wasp-85b , with an equivalent width upper limit of @xmath36m , corresponding to an abundance upper limit of @xmath488 , and implying an age of several myr @xcite .
stronger age constraints can be evaluated through a comparison to theoretical stellar models .
we use the isochrone placement method of @xcite , working in the @xmath489 $ ] parameter space suggested by @xcite .
we evaluate the age of each component using three sets of stellar models : the padova models of @xcite ; the yonsei - yale ( yy ) isochrones of @xcite ) , and models from the dartmouth stellar evolution database ( dsed ; @xcite ) .
we used independent measurements of ( derived from the stellar spectra ) and @xmath490 ( derived directly from the photometric lightcurves during the transit modelling process , see section[sec : orbit ] ) to derive the stellar ages at the central value of [ fe / h ] listed in table[tab : star_params ] , and at its @xmath491 limits .
the results are listed in table[tab : ages ] .
lll + method & + & wasp-85a & wasp-85b + & ( gyr ) & ( gyr ) + + + @xmath492 & @xmath493 & @xmath494 + + + padova & @xmath495 & @xmath496 + yy & @xmath497 & @xmath498 + dsed & @xmath499 & @xmath498 + garstec & @xmath500 & @xmath498 + + + @xmath446 & + & @xmath501 & @xmath502 + + + mamajek & @xmath503 & @xmath504 + + we also carry out a bayesian isochrone fitting process in @xmath505 $ ] parameter space , using the garstec models of @xcite , as computed by @xcite .
this new method is detailed in @xcite . using this method we find an age of @xmath500gyr for wasp-85a using @xmath8}=0.10 $ ] . , @xmath506 , and @xmath0gyr , while the green , dashed lines represent represent evolutionary models of @xmath507 , @xmath508 , and @xmath509@xmath510 , all computed for @xmath511_{\rm initial}=0.10 $ ] .
black dots mark individual steps in the markov chain ; we display only a random selection of @xmath512 data to improve clarity .
+ + , scaledwidth=48.0% ] in addition , we estimate the age of the two components using their measured and estimated rotation periods .
we use the gyrochronology formulation of @xcite , calculating the convective turnover timescale using table1 of @xcite and setting @xmath513d . using the measured rotation period ,
we derive a gyrochronological age of @xmath514gyr .
it is also possible to derive gyrochronological ages using the projected rotation velocity to derive an upper limit on the rotation period .
ages derived in this manner will also therefore be upper limits .
the rotation rate ( @xmath515d ) implied by the measured for wasp-85a gives an age of @xmath501gyr , in good agreement with this .
the implied rotation rate ( @xmath516d ) for wasp-85b gives an age of @xmath502gyr , which would seem to imply that the detected rotation period is that of wasp-85a .
however , given the contamination of the wasp photometry we can not be certain .
finally , we estimate the age for the two components using the stellar activity - age relation of @xcite . we find ages of @xmath517gyr and @xmath518gyr for stellar components a and b , respectively . the ages for wasp-85a that we derive using the four different isochrone analyses are consistent , and are also in agreement with the gyrochronology ages that we derive from the detected rotation period , and from .
the age that we determine using the chromospheric activity index is younger than many of the other estimates , but does agree with the padova isochrone age .
if we assume that the stars are gravitationally bound , as seems likely given the similar radial velocities that we measure for the two stars , and the fact that we see only a single ccf in the coralie and sophie spectrum , then it is encouraging that the results for the two stars agree .
we note though that the constraints on wasp-85b s age are very loose . to determine the planetary and orbital parameters , we carry out a simultaneous analysis of the harps rvs for wasp-85a , and the full suite of photometric observations .
this allows us to fully account for inter - parameter correlations , and properly characterise the uncertainties in the system parameters .
we elect not to include the rv data obtained using coralie and sophie owing to the aforementioned , unknown contamination caused by the presence of the companion star .
it is possible to correct the contaminated rvs through careful analysis of the ccfs , given that we have data for the two components separately from harps observations .
@xcite describe another means by which contaminated ccfs can be corrected that applies shifts to the spectrum and co - adds the results to the observed spectrum divided by a constant determined by the flux ratio .
given our access to uncontaminated harps data we elect not to perform a correction ; we leave this as a problem for a more detailed follow - up analysis of the system in the future .
we carry out our analysis using a modified version of the markov chain monte carlo ( mcmc ) code described in @xcite , @xcite , and @xcite . to recap , this models the rvs with a keplerian orbit , while the transits are fitted using the photometric model of @xcite .
we account for limb - darkening using a non - linear model with four components ; we derive wavelength appropriate coefficients for each photometric dataset by interpolating through the tables of @xcite at each mcmc step .
we also carry out a linear decorrelation in phase at each step to remove systematic trends in the photometric data ; owing to the presence of an additional trend in the lt / io : o lightcurves ( probably secondary extinction ) , we found it necessary to use a quadratic decor relation in phase for the two lightcurves obtained with that instrument .
we add an additional stellar `` jitter '' term of @xmath519to the formal rv uncertainties in order to obtain a reduced spectroscopic @xmath520 .
the following jump parameters are used as standard : the epoch of mid - transit , @xmath521 ; the orbital period , @xmath522 ; the full transit duration , @xmath523 ; the planet : star area ratio , @xmath524 ; the impact parameter for a circular orbit , @xmath525 ; ; [ fe / h ] , and the rv semi - amplitude , @xmath526 .
we apply priors on and [ fe / h ] using the spectroscopic values listed in table[tab : star_params ] , in all cases assuming a gaussian distribution .
steps are accepted using the metropolis - hastings algorithm , with a scale factor set to ensure an acceptance rate of roughly @xmath527percent . for wasp-85
, we add a number of additional jump parameters equal to the number of photometric datasets .
these are the third light contributions to each transit , @xmath528 , defined as the flux ratio of the two components @xmath529 , and allow us to account for the dilution induced by the presence of the binary companion .
we apply a bayesian prior on each of the @xmath528 , assuming a gaussian distribution ; for the observations in the sloan@xmath410 band we use the flux ratio of @xmath530 measured in the course of the io : o observations , while for the cousinsr and rg filters we scale this ratio assuming perfect blackbody emission for the two stars .
the third light contribution is added to the photometric transit model following the example of @xcite .
@xmath531 is strongly correlated with several other parameters , but we are able to model it in this fashion thanks to the measurement from io : o .
s. we plot the original data in grey , and the best fit transit model in red .
the lightcurves have been offset for clarity ; for the same reason , we omit the error bars .
the telescope , instrument , and date of observation are listed alongside each lightcurve .
a slewing problem is responsible for the gaps in the lightcurve obtained by trappist on 2013 - 03 - 31 .
+ + , scaledwidth=48.0% ] we test for the presence of a long - term trend in rv using the additional jump parameter @xmath532 , the rate of change in systemic velocity with time .
we find a significant improvement in the spectroscopic fit ( parameterised using @xmath533 ) when including such a trend , which we measure at @xmath534ms@xmath535yr@xmath535 .
however , we are sceptical that such a trend truly exists .
the timespan of the harps observations is only twelve days , covering just over @xmath536 orbits .
this is insufficient to truly constrain any trend that might be present in the systemic velocity .
moreover , the coralie and sophie data show no signs of any shift in @xmath537 over time . even with the dilution caused by wasp-85b
, a trend of @xmath534ms@xmath535yr@xmath535 would be easily identifiable in figure[fig : rvtime ] , as those data cover nearly @xmath538days , over five years , and the postulated trend is on the order of the detected semi - amplitude .
we therefore do not fit a trend in our final solution .
have been subtracted from each set of data . despite the strong trend identified in our fit ,
we see no evidence for such trend over the full timespan of our observations .
we conclude that the trend in @xmath539 arises owing to the very short baseline of the harps observations .
+ + , scaledwidth=48.0% ] to test for an eccentric orbit , we use additional jump parameters of @xmath540 and @xmath541 , where @xmath542 is the orbital eccentricity and @xmath543 is the argument of periastron . we find eccentricities of between @xmath544 and @xmath545 , depending on the inclusion or otherwise of a trend , and on whether the bayesian penalty on stellar radius was applied .
all are consistent with zero at @xmath546 .
we test the significance of these eccentricities using the f - test of @xcite to compare the fit to that of a circular orbit .
we find that the probability of the eccentricity being detected by chance varies between @xmath547 to @xmath548 , and therefore adopt a circular orbit in line with the advice in @xcite . once again
, the short timespan of the harps observations provides insufficient constraint even if the eccentricity was found to be significant .
one of the harps rv data has an uncertainty three times that of the other measurements , and a bisector span that is twice as large as the next greatest value .
we carried out runs omitting this datum to check that it was not biasing our results .
we found that removing this point made little difference to the significance of eccentricity , which we still find to favour a circular orbit but to be more significant in the absence of a long - term trend . removing
the datum led to only a small increase in the value of the fitted trend in @xmath537 .
we note though that the eccentricity of the orbit tended to increase with the removal of the highly uncertain harps measurement .
we attribute this to the spacing of the rv data in orbital phase , which leads to a fit dominated by the three data between phase @xmath459 and phase @xmath549 .
llll + parameter & symbol & value & units + + + projected rotation velocity & & @xmath445 & + effective temperature & & @xmath432 & k + metallicity & [ fe / h ] & @xmath450 & dex + flux ratio & @xmath531 & @xmath550 & cousins r filter + flux ratio & @xmath531 & @xmath550 & rg filter + flux ratio & @xmath531 & @xmath530 & sloan @xmath410 filter + + + orbital period & @xmath522 & @xmath551 & days + epoch of mid - transit & @xmath552 & @xmath553 & @xmath40 + transit duration & @xmath523 & @xmath554 & days + planet : star area ratio & @xmath524 & @xmath555 & + impact parameter & @xmath525 & @xmath556 & + rv semi - amplitude & @xmath557 & @xmath558 & ms@xmath535 + & @xmath559 & @xmath560 & ms@xmath535 + & @xmath561 & @xmath562 & ms@xmath535 + systemic velocity & @xmath539 & @xmath563 & + & @xmath564 & @xmath565 & + & @xmath566 & @xmath567 & + effective temperature & & @xmath568 & k + metallicity & [ fe / h ] & @xmath569 & dex + flux ratio & @xmath570 & @xmath571 & + & @xmath572 & @xmath573 & + & @xmath574 & @xmath575 & + + & @xmath576 & @xmath577 & + + & @xmath578 & @xmath575 & + & @xmath579 & @xmath580 & + & @xmath581 & @xmath550 & + & @xmath582 & @xmath575 & + & @xmath583 & @xmath584 & + + ingress / egress duration & @xmath585 & @xmath586 & days + orbital inclination & @xmath587 & @xmath588 & @xmath589 + orbital eccentricity & @xmath542 & @xmath459 ( adopted ) & + stellar mass & @xmath434 & @xmath590 & @xmath510 + stellar radius & @xmath437 & @xmath591 & @xmath440 + stellar surface gravity & & @xmath592 & ( cgs ) + stellar density & @xmath490 & @xmath593 & @xmath594 + planet mass & @xmath595 & @xmath596 & @xmath2 + planet radius & @xmath597 & @xmath598 & @xmath4 + planet surface gravity & @xmath599 & @xmath600 & ( cgs ) + planet density & @xmath601 & @xmath602 & @xmath603 + scaled semi - major axis & @xmath604 & @xmath605 & + semi - major axis & @xmath606 & @xmath607 & au + planet equilibrium temperature & @xmath608 & @xmath609 & k + + * notes : * @xmath610 : values of @xmath537 and @xmath526 for the coralie and sophie data were determined through a run using all three sets of rv data . this run did not significantly change the other model or derived parameters . the system parameters that we adopt are listed in table[tab : results ] , along with the prior values . in figure[fig : photometry ] we display the phase - folded photometric lightcurves , overlaid with the best fit transit model . note that the visible transit depth is shallower than the depth listed in table[tab : results ] owing to the effect of the third light contribution .
there is a notable difference in scatter between the two lt / io : o lightcurves .
this is likely to be a result of the comparison stars that were available for the data reduction process . for the lightcurve obtained on the night of 2013 - 01 - 12 there were two suitable comparison stars , which were fainter than than wasp-85 by factors of @xmath30 and @xmath611 . for the lightcurve obtained on the night of 2013
- 01 - 28 , only one comparison star was available , and it was four times fainter than our target . in figure[fig :
rvcurve ] we plot our rv data , overlaid with the best fit keplerian orbit model for harps only analysis .
we show all of our rv data to illustrate the effect of the dilution on the coralie and sophie observations .
the primary effect is a reduction in the semi - amplitude of the data ( see also table[tab : results ] ) , though there is also increased scatter compared to the harps observations .
we also show the residuals of the data compared to the plotted harps model .
the sinusoidal structure present in the residuals for the coralie and sophie data clearly indicates the reduced semi - amplitude .
this reduced semi - amplitude for the coralie and sophie data is expected given the cross - correlation method by which we derive their rv measurements .
wasp-85b is relatively constant in rv , and thus the peak of the ccf will be constant in velocity space .
the ccf for wasp-85a , however , will move in velocity space .
the close proximity of the two spares means that these ccfs merge to form a single , asymmetric ccf .
the peak velocity will be biased towards the stationary position of wasp-85b , effectively reducing the shift in velocity induced by the planet , and leading to a smaller semi - amplitude .
value for each set of data has been subtracted to allow comparison .
overplotted is our best fit orbital solution as derived from our mcmc analysis using the harps data only ( solid black line ) .
also shown are the keplerian solutions derived using @xmath526 for the coralie data ( dashed blue line ) and sophie data ( dotted red line ) .
the data for wasp-85b show strong variation , but it is not correlated with the phase of the planet s orbital solution . _ lower panel : _ radial velocity residuals as compared to the best fit model shown in the upper panel .
the effect of the dilution in reducing the semi - amplitude is clearly visible . + + , scaledwidth=48.0% ]
we test the effect of the dilution on the parameters that we determine by carrying out several different analyses .
we start by fitting the photometry and harps rvs using the same initial conditions , but without the additional third light term for the photometry .
this run finds a transit depth of @xmath612 , some @xmath14percent smaller than the depth when accounting for the contamination .
this is what we would expect given the flux ratio that we observe for the two stars .
this smaller transit depth leads to a stellar density of @xmath613@xmath594 , @xmath15percent smaller than the value we report in table[tab : results ] , and a planet radius of @xmath614@xmath4 , @xmath16percent smaller than the value we report . to characterise the effect of the third light dilution on the coralie and sophie rv data , we carry out an additional fit using the same initial conditions , but including all of our rv data .
we find that this makes little difference to the reported solution , but does provide additional useful information on the level of dilution . for this analysis
, we use a number of @xmath615 jump parameters equal to the number of datasets .
@xmath616 is significantly smaller than @xmath617 , which is in turn significantly smaller than @xmath557 .
this is as expected given the different instrument specifications , and the different levels of dilution that are expected in the three sets of data . in figure[fig : rvcurve ] we plot all rv data to illustrate this point ; note that the residuals are relative to the rv curve computed using @xmath557 .
as expected , the inclusion of the coralie and sophie data reduces the magnitude of the trend in systemic velocity to zero .
we also test the effect of the diluted rvs on the planet mass that we derive by carrying out runs using the coralie data or sophie data in place of the harps rvs .
when using the coralie rvs we find a planet mass of @xmath618@xmath2 , @xmath527percent smaller than the value we find using the harps data . with the sophie data , we find @xmath619@xmath2 , @xmath620percent smaller than the harps result .
these discrepancies match the differences that in @xmath615 that we find when fitting all of the rv data simultaneously , as expected .
we estimate the rv signature that would be induced in wasp-85a by the binary companion .
we calculate an expected rv semi - amplitude of @xmath621ms@xmath535 using the orbital period implied by the change in position angle .
the greatest negative rate of change of rv occurs at phase 0 ; adopting this as our initial condition , we calculate @xmath622ms@xmath535 over the course of one year .
wasp-85 is listed as a known binary in the washington double star catalogue ( wds ) .
we obtained the full set of wds measurements for the system , which stretch back to @xmath623 .
we also measured the position angle and angular separation of the companion star through careful analysis of eulercam observations taken on 2012 february 7th . in figure[fig :
pabinary ] we show both the binary angular separation and the binary position angle as a function of calendar date .
the change in binary separation appears to be roughly periodic , with a period of approximately @xmath624years .
however without error bars for the wds data , it is likely that this is just random scatter .
more high precision measurements of the separation will be needed to determine whether this variation is real ; the upcoming gaia mission will be of great help in this .
lll + date & position angle & angular separation + ( year ) & ( degrees ) & ( arc sec ) + + @xmath625 & @xmath626 & @xmath627 + @xmath628 & @xmath629 & @xmath630 + @xmath631 & @xmath632 & @xmath633 + @xmath634 & @xmath635 & @xmath636 + @xmath637 & @xmath638 & @xmath639 + @xmath640 & @xmath641 & @xmath642 + @xmath643 & @xmath644 & @xmath645 + @xmath646 & @xmath647 & @xmath648 + @xmath649 & @xmath650 & @xmath645 + @xmath651 & @xmath652 & @xmath653 + @xmath654 & @xmath655 & @xmath656 + @xmath657 & @xmath658 & @xmath659 + @xmath660 & @xmath635 & @xmath661 + @xmath662 & @xmath638 & @xmath663 + @xmath664 & @xmath665 & @xmath645 + @xmath666 & @xmath667 & @xmath633 + @xmath668 & @xmath669 & @xmath670 + @xmath671 & @xmath672 & @xmath627 + @xmath673 & @xmath674 & @xmath30 + @xmath675 & @xmath676 & @xmath677 + @xmath678 & @xmath679 & @xmath680 + + the position angle of the binary shows a clear , long - term , decreasing trend .
this is a clear indication of orbital motion for the binary , and suggests an orbital period of @xmath681years .
this change of position angle suggests that the planet is misaligned from the plane of the binary . using the distance listed in table[tab : star_params ] , the mean separation of @xmath656corresponds to a distance between the stars of @xmath414au . using the masses that we estimate for the two components from our spectral analysis , and assuming a circular orbit , this indicates a period of @xmath682years , far less than the @xmath683years suggested by the position angle change .
the reverse transformation , from @xmath684years , gives a binary distance of @xmath685au ; again , this is discrepant with the separation derived value .
this mismatch suggest that the binary orbit is inclined relative to our line of sight by @xmath686 , again indicating a misalignment with the planet s orbital plane .
the minimum separation implies an orbital distance of @xmath687au . _
lower panel : _ binary position angle as a function of calendar date .
there is a clear , long - term trend for decreasing position angle .
this suggests an orbital period for the binary of @xmath681years .
+ in both panels , black circles indicate data obtained from the washington double star database , while blue triangles indicate data derived from eulercam observations . + + ,
scaledwidth=48.0% ] the presence of the binary companion at @xmath688au from the planet host star suggests that the protoplanetary disc was likely truncated , limiting the quantity of material available for planet formation .
exploring this possibility is beyond the scope of this paper , but could be an interesting subject for future work . comparing the caiih+k indices for the two stars to the sample of @xcite , we find that both stars fall within the ` very active stars ' region of figure4 in @xcite .
this is confirmed by figure10 of @xcite and figure6 of @xcite , where in both cases the two stars fall within the secondary , ` active ' peak in the @xmath452 distribution .
comparison to the sample of @xcite shows that both stars fall in the ` active ' class of stars . comparing to the sample of planet hosting stars examined by @xcite
, we see that wasp-85a is more active than all but one of the stars considered .
only corot-2 is more active
. it may be that we have simply observed the system at the peak of the activity cycle , leading to an apparently greater level of activity .
however , if we consider the solar cycle then the typical variation is @xmath689dex ; converting the solar caiih+k values presented in @xcite indicates values of @xmath690 and @xmath691 at solar minimum and maximum , respectively .
measurements of activity during the maunder minimum @xcite indicate @xmath692 during that particularly inactive period in the sun s life , giving a pessimistic variation from maximum of @xmath693dex . if we apply this to wasp-85a , then even at stellar ` minimum ' it will be more active than the sun at solar maximum , and will still be classified in the ` active ' class of @xcite .
it therefore seems that these stars are unusually active for solar - type stars .
the amplitude of rotational variability in k - dwarfs can reach a few percent for periods @xmath694days @xcite .
it is therefore possible that the modulation observed in the wasp lightcurve originates from wasp-85b .
the negative correlation that we observe between the harps bisector span and rv measurements for wasp-85b suggest that this might be the case .
though it is fainter , the difference in magnitude is such that the companion contributes significant light ( as can be seen from the flux ratio of the two stars ) .
moreover the projected rotational velocities of the two stars are similar , and both are compatible with the measured rotation period .
this degeneracy could be broken by careful , high resolution photometry of the two stars to check their individual rotation periods , or by measurement of the stellar inclinations .
if either star is significantly inclined relative to the line of sight , then its true rotation period would be significantly shorter , and thus incompatible with the period determined from the wasp photometry . during our campaign of follow - up observations of wasp-85ab , a search of the proposed k2 fields revealed that the planetary system was visible in the campaign1 field .
we proposed short cadence observations of wasp-85 under k2 proposal go1041 , and the system was selected for observation by k2 .
four other k2 campaign1 proposals ( go1032 , go1054 , go1059 , and go1005 ) that included wasp-85 were also selected , three for long cadence observations and one for short cadence observations of the double star . on the release of the campaign1 data
, we downloaded the long cadence data for wasp-85 ( short cadence data were not immediately available ) , extracted the photometric data , and corrected it for the systematic effect of pointing drift .
this has been identified as the greatest source of error in the k2 lightcurves @xcite .
our detrending method is similar to that developed by @xcite , and has been previously described in @xcite .
briefly , we use the centroid positions to create a 2d surface of raw flux as a function of the centroid s position on the ccd .
we then bin this data evenly by both row and column , and linearly interpolate the median flux from each bin to create a variability map .
we divide the raw photometry by this surface map to create a corrected flux measurement that is decor related from the spacecraft s pointing .
the planetary transit signatures are clearly visible , as is variation arising from stellar activity .
this stellar activity signal is of lower amplitude than the planetary transits .
the system was also observed in short cadence mode as part of k2 campaign1 , under our proposal go1041 ; this data was not available at time of publication .
_ lower panel : _ the k2 long cadence lightcurve , phase folded on the best fit ephemeris from table[tab : results ] .
we show a small window around phase0 to focus on the planetary transit .
stellar activity has been corrected for using a simple median filter .
+ uncertainties are plotted in both panels , but are smaller than the size of the data points.,scaledwidth=48.0% ] the upper panel of figure[fig : k2 ] shows the detrended flux as a function of time .
the planetary transits are clearly visible where they extend below the stellar lightcurve . also apparent
is the signature of stellar activity ; the variations caused by the activity of one or both components of wasp-85ab appears to be aperiodic at first glance , but this may result from star spot evolution .
an initial lomb - scargle periodogram analysis of the lightcurve implies a period of @xmath695days for the variability , roughly commensurate with half of the rotation period determined from the wasp data .
the amplitude of the activity signal is approximately half the transit depth .
the lower panel of figure[fig : k2 ] shows the detrended k2 data phase folded on the best fit ephemeris from table[tab : results ] .
we show only a small window around phase0 to focus on the planetary transit , and have corrected for the stellar activity using a simple median filter with a window size of @xmath696 times the length of the time series .
this simple correction for activity leads to slightly increased scatter between contact points two and three of the transit .
the transit depth is approximately @xmath68magnitudes , less than the transit depth listed in table[tab : results ] .
this is to be expected given the large pixel scale of the kepler spacecraft s ccd , which means that the stellar components of the system will be blended together as was the case for our other photometry .
the depth discrepancy is as expected given the kepler passband , the peak transmission of which roughly coincides with the peak transmission of the johnsonr filter @xcite .
analysis of the k2 data is ongoing , and incorporates both the long cadence data presented here and the short cadence data that was also collected by k2 .
the results of this analysis will be presented in a follow - up paper .
we have presented the discovery of wasp-85ab , a hot jupiter planet orbiting the brighter component of bd+07@xmath242474 every @xmath0days .
the planet has a mass of @xmath22@xmath2 and a radius of @xmath23@xmath4 .
the host star is similar to the sun , but has super - solar metallicity , while the companion is a cooler k - dwarf of similar magnitude .
the caiih+k activity indices for the two stars indicate that both are strongly active , particularly when compared to other hot jupiter hosts .
our photometric observations of wasp-85 are contain light from both binary components , necessitating the addition of a third light component to the standard transit model , and the modification of our standard wasp analysis routines .
spectroscopic observations with coralie and sophie are also diluted by the companion , so we use only radial velocity data obtained with harps to perform our modelling .
we find that the primary effect of the dilution is to decrease the radial velocity semi - amplitude for the affected datasets . if we fail to account for the contamination of the photometry , the derived planet radius decreases by @xmath16percent .
the binary position angle shows a clear , long - term , negative trend that suggests an orbit of @xmath681years .
this is at odds with the period derived from the mean binary angular separation , indicating that the binary is likely inclined relative to our line of sight by @xmath686 .
the planet s orbit is therefore misaligned with the binary plane .
bd+07@xmath242474 has been observed at short cadence as part of k2 campaign1 .
we have presented the k2 long cadence lightcurve for the system , which clearly shows the variability resulting from stellar activity , with the planetary transits superposed .
k2 is unable to distinguish the binary components .
analysis of this new data is ongoing , and will be presented in a follow - up paper .
the wasp consortium consists of representatives from the universities of keele , leicester , the open university , queens university belfast and st andrews , along with the isaac newton group ( la palma ) and the instituto de astrofisica de canarias ( tenerife ) .
wasp - south is hosted by the saao and superwasp - north by the isaac newton group and the instituto de astrofisica de canarias ; we gratefully acknowledge their ongoing support and assistance .
the superwasp and wasp - s cameras are operated with funds made available from consortium universities and the stfc .
trappist is funded by the belgian fund for scientific research ( fond national de la recherche scientifique , fnrs ) under the grant frfc 2.5.594.09.f , with the participation of the swiss national science foundation ( snf ) .
the liverpool telescope is operated on the island of la palma by liverpool john moores university in the spanish observatorio del roque de los muchachos of the instituto de astrofisica de canarias with financial support from the uk science and technology facilities council .
m. g. and e. j. are fnrs research associates .
l. d. is an fnrs / fria doctoral fellow .
ahmjt is a swiss national science foundation fellow under grant number p300p2 - 147773 .
this research has made use of nasa s astrophysics data system bibliographic services , the arxiv preprint service hosted by cornell university , the washington double star catalog maintained at the u.s .
naval observatory , and the vizier catalogue access tool , cds , strasbourg , france .
the original description of the vizier service was published in a&as 143 , 23 .
queloz d. , mayor m. , naef d. , santos n. , udry s. , burnet m. , confino b. , 2000 , in bergeron j. , renzini a. , eds , the vlt opening symposium : from extrasolar planets to cosmology , springer - verlag , berlin , p. 548
santerne a. , moutou c. , bouchy f. , 2011 , in ` the astrophysics of planetary systems : formation , structure , and dynamical evolution ' , proceedings of iau symposium 276 , eds .
sozzetti a. , lattanzi m. g. , boss a. p. , p. 549 | we report the discovery of the transiting hot jupiter exoplanet wasp-85ab . using a combined analysis of spectroscopic and photometric data , we determine that the planet orbits its host star every @xmath0days , and has a mass of @xmath1@xmath2 and a radius of @xmath3@xmath4 .
the host star is of g5 spectral type , with magnitude @xmath5 , and lies @xmath6pc distant .
we find stellar parameters of @xmath7k , super - solar metallicity ( @xmath8}=0.08\pm0.10 $ ] ) , @xmath9@xmath10 and @xmath11@xmath12 . the system has a k - dwarf binary companion , wasp-85b , at a separation of @xmath13 .
the close proximity of this companion leads to contamination of our photometry , decreasing the apparent transit depth that we account for during our analysis . without this correction , we find the depth to be @xmath14percent smaller , the stellar density to be @xmath15percent smaller , and the planet radius to be @xmath16percent smaller than the true value .
many of our radial velocity observations are also contaminated ; these are disregarded when analysing the system in favour of the uncontaminated harps observations , as they have reduced semi - amplitudes that lead to underestimated planetary masses .
we find a long - term trend in the binary position angle , indicating a misalignment between the binary and orbital planes .
wasp observations of the system show variability with a period of @xmath17days , indicative of rotational modulation caused by stellar activity .
analysis of the caiih+k lines shows strong emission that implies that both binary components are strongly active .
we find that the system is likely to be less than a few gyr old .
wasp-85 lies in the field of view of k2 campaign1 .
long cadence observations of the planet clearly show the planetary transits , along with the signature of stellar variability .
analysis of the k2 data , both long and short cadence , is ongoing . |
it is well known that the average size @xmath11 of an isolated flexible uncharged polymer chain in good solvent is asymptotically proportional to @xmath12 , where @xmath2 is the degree of polymerization , and @xmath13 @xcite .
this law holds for any measure of the chain size , the most popular of which are the mean square end to end distance , @xmath14 the mean square radius of gyration , @xmath15 and the hydrodynamic radius @xmath16 in these equations , we have assumed that the chain is composed of @xmath2 monomers ( i. e. @xmath17 bonds ) at positions @xmath18 , @xmath19 , and @xmath20 .
experimentally , the gyration radius is determined from small angle scattering experiments ; for small wave numbers @xmath21 the single chain static structure factor behaves like @xcite @xmath22 \right > \nonumber \\ & = & n \left [ 1 - \frac{q^2}{3 } \left < r_g^2 \right > + o(q^4 ) \right ] . \end{aligned}\ ] ] conversely , the hydrodynamic radius is determined via small angle dynamic light scattering experiments , where the dynamic structure factor @xmath23 \right>\ ] ] is measured . in the small @xmath21 limit , it decays like @xmath24 where @xmath25 is the chain diffusion constant , which , within the framework of kirkwood zimm theory @xcite is related to @xmath0 via @xmath26 where @xmath27 is the monomer diffusion constant ( usually this contribution is neglected ) , @xmath28 is boltzmann s constant , @xmath29 the temperature and @xmath30 the solvent viscosity . in principle , it is possible to obtain @xmath0 , as a static quantity , also from purely static scattering , as is seen from the relation @xmath31 but for technical reasons , this has so far not been applied in experiments . when analyzing data for the chain size , one has to take into account that the law @xmath32 holds only in the asymptotic limit @xmath33 , while for finite chain lengths deviations occur .
this is particularly important for computer simulations , where data with high statistical accuracy can be obtained .
for this reason , corrections to scaling have been worked out in great detail , and exploited in high resolution numerical studies , for the end to
end distance and the gyration radius , where the relation @xmath34 ( and analogously for @xmath7 ) holds @xcite . here
@xmath35 and @xmath36 are non universal amplitudes , while @xmath37 is a universal correction to scaling exponent , whose value is difficult to determine beyond the accuracy @xmath38 ( @xmath39 according to ref .
, @xmath40 according to ref . ) .
the omitted terms are further powers @xmath41 , @xmath42 , , as well as @xmath43 , @xmath44 , ( i .
e. there are further larger correction to scaling exponents ) , plus so called `` analytic '' terms @xmath45 , @xmath46 ,
the important point to notice is that the `` analytic '' corrections will arise even for a gaussian chain , and are due to the fact that the chain consists of a finite number of beads .
this will be demonstrated explicitly in sec .
[ sec : hydradgaussian ] .
conversely , the `` non
analytic '' corrections are due to the fact that , in the language of renormalization group ( rg ) theory , the chain s hamiltonian is not identical to the fixed point hamiltonian .
the exponent @xmath37 is related to the largest sub leading eigenvalue of the rg transformation at the fixed point . in first order @xmath47 expansion
its value is @xcite @xmath48 with @xmath49 and @xmath50 , where @xmath51 and @xmath52 is the spatial dimension .
higher order calculations @xcite have resulted in @xmath53 .
we adopt here the convention ( which we view as quite natural ) to distinguish the terms by their different origins , and call correction terms `` analytic '' corrections if they are present even in the gaussian limit , while we call `` non analytic '' corrections those terms which occur exclusively for excluded volume chains . as we will see in sec .
[ sec : hydradgoodsolv ] , `` analytic '' corrections defined in this way do not necessarily imply integer powers of @xmath2 . as for the hydrodynamic radius of self avoiding walks ( saws ) , there is no high resolution numerical study available , and corrections to scaling have not yet been dealt with systematically .
this is somewhat unfortunate , since the corrections are unusually large for @xmath0 , and of experimental relevance . for a good solvent chain ,
one expects again @xmath54 etc .
terms , plus `` analytic '' corrections .
it is the main purpose of the present paper to show that the leading order term of these latter corrections is now given by @xmath55 where @xmath36 is usually positive .
we will show in secs . [
sec : hydradgoodsolv ] and [ sec : altderiv ] that this form is a straightforward consequence of discretizing the chain into beads . as @xmath56 , this will , for long chains , ultimately dominate over the @xmath54 term , where the exponent is ( according to refs . , and ) slightly larger .
nevertheless , the exponents are so close that in most cases one will observe contributions from both terms . on the other hand
, it is a well known empirical fact that for many experimental systems , as well as for most computer models , the corrections to scaling of the gyration radius are quite weak , such that the @xmath54 term should have a rather small amplitude .
one could therefore expect that the corresponding amplitude of the @xmath54 contribution in @xmath0 is also quite small .
then the most likely candidate for explaining the experimental and numerical observation that @xmath0 is usually subject to very large corrections to scaling @xcite would actually be the `` analytic '' @xmath57 term . for a _
gaussian _ chain
we are able to solve the problem exactly , see sec .
[ sec : hydradgaussian ] : @xmath58 with @xmath59 ( here @xmath60 is riemann s zeta function ) , and @xmath61 denoting the root mean square bond length . as @xmath62 for a gaussian chain ,
this form is consistent with eq .
[ eq : analcorrhydrad ] .
the difficulties in observing the asymptotic @xmath12 scaling of @xmath0 have a long history .
adam and delsanti @xcite performed dynamic light scattering experiments and found an effective power law @xmath63 .
this is quite typical , and has been found in many other experiments , too @xcite , although an exponent of @xmath64 @xcite has been reported as well .
a reduction of the effective exponent is indeed expected , as seen from eq .
[ eq : analcorrhydrad ] , and is also observed in brownian dynamics simulations @xcite . as a caveat , note that a scattering experiment does not measure @xmath0 , but rather the diffusion constant @xmath25 .
this quantity has an additional @xmath65 contribution ( eq .
[ eq : kirkwoodformula ] ) , which is of the same order as the leading correction of eq .
[ eq : analcorrhydrad ]
. therefore , the corrections in @xmath25 are weaker than those in @xmath0 .
nevertheless , the @xmath65 term is typically not large enough to fully compensate the corrections in @xmath0 .
this is easily seen for the gaussian case from eqs .
[ eq : kirkwoodformula ] and [ eq : gaussmainz ] : the monomer diffusion constant @xmath27 can be written as @xmath66 , which defines a monomer stokes radius @xmath67 .
thus @xmath68 since @xmath67 should be of the order of the bond length @xmath61 , one sees that a large @xmath45 contribution remains . a first attempt to explain the experimental observation is due to weill and des cloizeaux @xcite .
they conjectured that @xmath69 is due to non perfect solvent quality , and a crossover between good solvent behavior ( @xmath70 , large length scales ) and @xmath8 solvent ( @xmath71 , small length scales ) .
in particular , they pointed out that the averaging over @xmath72 assigns a very large statistical weight to the small distances .
although this latter argument is true , and generally accepted as the basic origin for the slow convergence of @xmath0 , the explanation in terms of solvent quality has turned out to be incorrect . in ref .
it was demonstrated that @xmath0 should _ not _ be much more susceptible to solvent quality effects than @xmath73 or @xmath7 the enhanced sensitivity of @xmath0 to the small distances is balanced by the fact that @xmath73 and @xmath7 are more sensitive to the decreased swelling of the chain near its ends : a saw is _ inhomogeneous _ , i. e. @xmath74 depends on the position of the @xmath75 bond on the chain , and is systematically larger in the interior , as has been shown both numerically @xcite and analytically @xcite .
furthermore , schfer and baumgrtner @xcite performed a detailed rg calculation and predicted in one loop order for the universal amplitude ratio @xmath76 here @xmath77 is the exact random walk ( rw ) value .
other rg studies resulted in @xmath78 @xcite ( this value was later revised to 1.51 , see ref . ) and @xmath79 @xcite , while a semi empirical relation based on fitting the distribution function of internal distances to light scattering data yields @xmath80 @xcite .
a value of @xmath81 was also found in brownian dynamics simulations @xcite . from eq .
[ eq : analcorrhydrad ] it is clear that @xmath82 should be subject to an @xmath4 correction for finite chain length ; nevertheless , experiments have so far not reported a systematic dependence on molecular weight .
typically , values around @xmath83 @xcite , or @xmath84 / @xmath85 for different solvents @xcite are found in the good solvent regime . in view of the inaccuracies of the experiments ( @xmath7 typically has an error of @xmath86 @xcite ) the inability to observe a systematic behavior in @xmath2 is not very surprising . in order to contribute to the resolution of these questions , we have performed computer simulations of very different models of polymer chains , both for saws and for @xmath8 chains , and calculated @xmath7 and @xmath0 , as outlined in sec .
[ sec : numres ] . to provide
a complete and well converged data set represents the second main goal of our paper . to our knowledge , our results are the most accurate data obtained on @xmath0 so far .
concerning @xmath7 , however , our data are less accurate than those of li _ et al . _
@xcite , which we still view as the most precise numerical study on the saw problem so far .
therefore we have taken their values for the exponents @xmath87 and @xmath37 for our fits .
we find @xmath88 for good solvent chains , in very good agreement with ref . .
( note that our error estimate is probably overly optimistic , since it only includes statistical errors and completely neglects systematic errors .
) theoretical and numerical investigations on corrections to scaling in @xmath0 have first focused on the rw case .
the work by guttman _
@xcite showed by analytical calculation that a gaussian chain should obey eq .
[ eq : gaussmainz ] .
the prefactor of the correction was first @xcite determined only approximately , @xmath89 , while later @xcite it was given exactly in terms of an integral .
furthermore , monte carlo ( mc ) simulations of lattice chains at the @xmath8 point revealed that in this case the ratio @xmath6 apparently does _ not _ converge to its gaussian value @xmath10 , but rather to roughly @xmath90 .
our simulations ( see sec . [
sec : numres ] ) find a similar behavior ( @xmath91 ) .
we believe that this can be explained qualitatively from rg arguments @xcite as follows : the asymptotic behavior is expected to be governed by the gaussian fixed point , and thus @xmath92 , as a universal amplitude ratio , is expected to assume the gaussian value .
however , the numerical extrapolation will only produce this value if all relevant correction terms , i. e. the analytic @xmath93 term , plus the non analytic corrections , are consistently taken into account . neither the data analysis by guttman _ et al . _ @xcite , nor ours , fulfill this requirement , as both just fit to eq .
[ eq : analcorrhydrad ] with @xmath62 , and thus are expected to produce substantial systematic errors in @xmath92 .
to do this in a better way is practically impossible , since ( i ) our @xmath8 data have insufficient statistical accuracy to allow for additional fit parameters , ( ii ) the precise form of the correction terms is unknown for @xmath0 ( in contrast to @xmath73 and @xmath7 , for which the leading order terms have been calculated by tricritical field theory @xcite , with the interesting feature that they are universal ) , and ( iii ) the non analytic corrections vary extremely slowly ( logarithmically ) with @xmath2 , such that either one would need unrealistically long chains to ensure dominance of the leading orders , or an expansion up to unrealistically high order .
these problems have been elucidated in quite some detail for @xmath73 and @xmath7 @xcite , explaining previous difficulties in the interpretation of highly accurate mc data on @xmath8 chains @xcite . in this context
, it should be mentioned that experiments @xcite typically find a value of @xmath94 , i. e. a similar reduction as in the good solvent case .
later , mc data were taken of excluded volume ( ev ) chains with saw statistics .
schfer and baumgrtner @xcite used chains of up to @xmath95 monomers , with an ev strength particularly close to the saw fixed point , such that poor solvent effects can be ruled out .
the inhomogeneous swelling was demonstrated , and the @xmath0 data were fitted with eq .
[ eq : analcorrhydrad ] .
this was done with an empirical correction to scaling exponent of @xmath96 instead of @xmath97 .
the same evidence was shown in the simulation data by batoulis and kremer @xcite of chains of length of up to @xmath98 .
ladd and frenkel @xcite simulated chains of length of up to @xmath99 and were able to describe their @xmath0 data via eq .
[ eq : analcorrhydrad ] , with @xmath100 and @xmath101 , but without detailed justification of their use of the correct @xmath97 exponent .
schfer and baumgrtner @xcite concluded from both their analytical studies and their simulation data that not the solvent quality , but rather the chain s microstructure is responsible for the slow convergence . our reasoning ( secs .
[ sec : hydradgaussian][sec : altderiv ] ) , which is similar to the one by guttman _
@xcite , exactly supports this picture : the corrections are due to the fact that the chain is discretized into beads , or , in other words , to the fact that there is a lower length scale cutoff for the frictional properties . however , the notion of `` stiffness '' , which is often used in this context @xcite , is , in our view , somewhat misleading : as outlined in sec .
[ sec : stiffness ] , we expect a large local chain stiffness to _ decrease _ the correction until it ultimately even changes its sign . the same conclusion has been found by akcasu and guttman @xcite for stiff chains without excluded volume
. in the context of dynamic light scattering of dilute polymer solutions there is yet another unresolved puzzle .
as akcasu _
et al . _ have shown @xcite , the initial decay rate of the dynamic structure factor , @xmath102 can be written as @xmath103 where @xmath104 is the diffusion tensor .
equation [ eq : omegaofqakcasu ] is a rigorous result , the only assumption being that the chain dynamics can be described by kirkwood s diffusion equation @xcite .
usually , @xmath104 is taken as the oseen tensor , @xmath105 where @xmath106 denotes the tensor product of the unit vector in @xmath107 direction with itself . in this case , eq .
[ eq : omegaofqakcasu ] is just the @xmath108 generalization of eq .
[ eq : kirkwoodformula ] .
it can then be shown @xcite that for @xmath21 in the scaling regime @xmath109 ( @xmath61 denoting the bond length ) , or , strictly spoken , in the limit @xmath110 , @xmath111 , the relation @xmath112 holds , where the numerical constant @xmath113 only depends on chain statistics : @xmath114 for rw statistics ( @xmath62 ) and @xmath115 for saws ( @xmath116 ) .
this has been tested by light scattering experiments both for good solvents @xcite and for @xmath8 solvents @xcite . in both cases
the relation is verified with reasonable accuracy , but with a prefactor @xmath113 which is systematically smaller than the theoretical prediction .
the reasons for this shift are not clear ; an attempt by a generalized theory which introduces draining @xcite so far had only limited success @xcite . in sec .
[ sec : initdecay ] we show that the deviation can partly be explained by the fact that in reality neither @xmath117 nor @xmath118 holds . taking these nonidealities crudely into account
, we find a shift in the same direction , which is however smaller than the experimental one .
nevertheless , we believe that this third main result is of direct relevance for the analysis of experimental data .
there are also some indications from molecular dynamics simulations @xcite that the description in terms of the kirkwood theory is insufficient on these length and time scales .
for a gaussian chain with root mean square bond length @xmath61 , we have @xmath119 and @xmath120 and hence @xmath121 and @xmath122 where we have used elementary summation formulae . for the hydrodynamic radius ,
we find analogously @xmath123 according to the euler
maclaurin formula ( see appendix [ sec : euler ] ) , eq .
[ eq : eulermaclaurinpower ] , the sums can be expanded as @xmath124 hence , @xmath125 and @xmath126 which is the result anticipated in eq . [
eq : gaussmainz ] . for a linear
saw , the main difficulty is the fact that , unlike for a rw , @xmath127 and @xmath128 do _ not _ @xcite just depend on @xmath129 , but rather on the positions relative to the ends as well . in order to obtain the leading order analytic corrections due to discretization , we can restrict the discussion to the leading order scale invariant behavior , i. e. we can assume that the saw is strictly scale invariant with the exponent @xmath87 , with no non analytic corrections . if we would include the latter , they would just generate further additive terms in our expressions . in
what follows , we therefore omit them , for the sake of simplified notation , but keep in mind that they have to be added at the end in order to obtain the full expressions .
we thus assume the relations @xmath130 where we have introduced the notation @xmath131 the definitions of @xmath7 and @xmath0 lead us to study the sum @xmath132 for @xmath133 and @xmath134 , respectively , by means of the euler
maclaurin expansion of appendix [ sec : euler ] . treating the inner sum first
, we find @xmath135 with the formal expansion @xmath136 since @xmath137 vanishes .
note also that the constant in eq .
[ eq : eulersaw1 ] does not depend on @xmath138 , hence @xmath139 inserting eq .
[ eq : eulersaw2 ] , we find @xmath140 after transformation to the reduced variables @xmath141 and @xmath142 , and exploiting the scaling behavior of @xmath143 , it is possible to determine the order of each term . for the gyration radius , we find @xmath144 , @xmath145 , @xmath146 , @xmath147 , @xmath148 , respectively , for the five terms in order of their appearance .
conversely , for @xmath149 the different scaling behavior of @xmath143 implies @xmath150 , @xmath151 , @xmath152 , @xmath147 , @xmath148 for the corresponding orders . for @xmath153
the leading order is @xmath144 , while the next sub
leading order is @xmath146 , resulting in a leading order correction of @xmath154 . for @xmath155 ,
the leading order is @xmath150 , followed by the @xmath147 term .
thus the correction to scaling for @xmath7 has order @xmath156 .
this proves eq .
[ eq : analcorrhydrad ] .
of course , this consideration does not prove that the amplitude @xmath36 in eq .
[ eq : analcorrhydrad ] is positive ; however , this is expected from the result for gaussian chains .
equation [ eq : analcorrhydrad ] can also be derived in a more heuristic way , starting from eq .
[ eq : rhfromsofq ] .
figure [ fig : sofq ] shows the typical behavior of @xmath157 : for wave numbers @xmath21 with @xmath109 the structure factor exhibits a power law decay @xmath158 which indicates the chain s fractal geometry , while for larger @xmath21 it oscillates around unity .
we therefore can introduce a cutoff wavenumber @xmath159 from which on there is no further contribution to the integral , i. e. @xmath159 is the smallest of all the @xmath160 s with the property @xmath161 . hence , @xmath162 it is physically clear that for a flexible chain @xmath159 must be roughly @xmath163 , apart from a numerical prefactor of order unity .
moreover , the fractal @xmath158 decay of @xmath157 roughly extends up to @xmath159 , at which point @xmath164 is reached .
we now introduce a modified structure factor @xmath165 , which is identical to @xmath157 up to @xmath166 , but extends the @xmath167 decay up to @xmath168 . in this latter regime
, we have @xmath169 where @xmath170 is a numerical prefactor of order unity .
therefore we can write @xmath171 evaluating the second integral , and writing @xmath165 in scaling form , @xmath172 ( here we have again assumed strict scale invariance , i. e. absence of non analytic corrections to scaling , for the same reason as outlined at the beginning of the previous subsection ) , one finds @xmath173 i. e. again a negative correction of order @xmath156 .
the advantage of the approach of the previous subsection is that it can be easily generalized to study the influence of local structure , since it is well known how this is reflected in @xmath157 . for a locally stiff chain with a persistence length large compared to the bond length @xmath61 ,
one expects that @xmath159 is roughly unchanged with respect to the flexible case .
however , the @xmath167 decay does no longer extend down to @xmath174 , but only to @xmath175 , where @xmath176 is a crossover wave number , whose inverse is a typical length scale below which stiffness effects are important . with @xmath177 being again the continuation of the @xmath167 decay up to @xmath178 , we have @xmath179 we now assume @xmath180 for @xmath181 , and @xmath182 for @xmath183 . here ,
@xmath170 and @xmath184 denote prefactors of order unity , and the @xmath185 decay results from the local stretching . evaluating the integrals , and using eq .
[ eq : strufacscalingform ] , one thus finds @xmath186 in order to compare with the flexible case , we still have to take into account that stiffness tends to increase the gyration radius , by roughly a factor of @xmath187 : @xmath188 where @xmath189 denotes the gyration radius in the flexible case .
the prefactor of the correction term hence depends on the stiffness parameter @xmath190 in a non trivial way ; for small @xmath190 both an increase and a decrease are possible , depending on the parameters . for sufficiently large
stiffness one always obtains a decrease of the correction , and ultimately even a change of its sign . in this subsection
, we are concerned with the initial decay rate @xmath191 , see eq .
[ eq : omegaofqakcasu ] . splitting the sum in the numerator into diagonal and off diagonal terms
, one finds @xmath192 following refs .
, we use the fourier representation of the oseen tensor for the off
diagonal elements , @xmath193 to find @xmath194 we now focus attention on the dimensionless quantity @xmath195 where we again have expressed the monomer diffusion constant @xmath27 in terms of a stokes radius @xmath67 . after transforming to the dimensionless integration variable @xmath196 and performing the angular integration , one has @xcite
@xmath197 with @xmath198 this function can be expanded as @xmath199 for @xmath200 , and @xmath201 for @xmath202 . in order to make further progress
, we have to specify the structure factor @xmath157 .
this shall be done by the most simplistic model which takes into account both finite bead size and finite chain length ( see also fig . [
fig : sofq ] ) : @xmath203 we now calculate @xmath204 in the scaling regime @xmath205 . defining the @xmath206 values where @xmath207 changes its behavior as @xmath208 we can write @xmath209 , @xmath210 , @xmath211 ; hence @xmath212 since @xmath213 and @xmath214 , we can write @xmath215 where we have taken just the leading order terms of the expansions of @xmath216 ; this results in @xmath217 in the asymptotic limit @xmath111 , i. e. @xmath218 , and @xmath219 , i. e. @xmath220 , this obviously converges to the asymptotic value @xmath221 focusing now on the correction , i. e. @xmath222 , we find @xmath223 taking again the leading order terms for the remaining integrals results in @xmath224 one thus sees that _ both _ finite chain length _ and _ finite bead size have the tendency to decrease @xmath113 , as observed in the experiments .
the latter effect is clearly more important , as @xmath225 enters only via a relatively high power .
further insight is gained by numerical evaluation of the shift for reasonable parameter values .
tsunashima _ et al . _
@xcite performed their experiments with polyisoprene chains of size @xmath226 .
typical scattering wavenumbers in their plateau regime were given by @xmath227 ; the experimental observation in this regime was @xmath228 , i. e. a shift of @xmath229 .
in what follows , we consider the value @xmath230 .
thus @xmath231 where we have estimated the monomer size @xmath67 as @xmath232 @xcite .
inserting this into eq .
[ eq : shiftresult ] , we find for the @xmath233 contribution a value of @xmath234 , i. e. one order of magnitude smaller than the experimental value .
it is not completely clear if a more thorough treatment of the integral would fully account for the deviation ; our guess is that it would probably not .
molecular dynamics data @xcite seem to rather indicate that for typical systems ( i. e. on not yet asymptotic length scales ) the coupling between polymer and solvent is more complex than the simple kirkwood description .
nevertheless , we consider our result as important for the interpretation of experimental data : there is obviously a substantial contribution to @xmath235 which stems from the finite bead size , and which is only weakly @xmath21dependent .
a plateau like shape of @xmath204 alone apparently does not guarantee asymptotic behavior .
clearly more work has to be done to fully resolve the puzzle , but we believe our considerations show that theories which neglect the influence of finite bead size ( and , to a lesser degree , of finite chain length ) are simply not accurate enough to describe experimental data even of rather long chains .
in our numerical studies , we have used three different polymer chain models , which we will denote as model a , b , and c. _ model a _ is a bead spring model in the continuum .
@xmath2 monomers are connected via an anharmonic ( `` finitely extensible nonlinear elastic '' ) spring potential , @xmath236 & \hspace{0.5 cm } r < r_0\\ \infty & \hspace{0.5 cm } r \geq r_0 \end{array } \right . , \ ] ] where we use the standard parameters @xcite @xmath237 , @xmath238 in dimensionless units
. between all monomers there is an additional non bonded potential @xmath239 - \lambda , \hspace{0.2 cm } r \le 2^{1/6 } \\
\frac{1}{2 } \lambda \left [ \cos ( \alpha r^2 + \beta ) - 1 \right ] , \hspace{0.2 cm } 2^{1/6 } \le r \le 1.5 \\ \\ 0 , \hspace{0.2 cm } r \ge 1.5 , \end{array}\right.\ ] ] where @xmath170 and @xmath184 are determined as the solutions of the linear set of equations @xmath240 i. e. @xmath241 and @xmath242 .
this potential has originally been constructed to simulate amphiphilic systems @xcite .
the parameter @xmath243 serves to control the strength of the attractive interaction and is varied instead of the temperature , which is fixed at @xmath244 . for sufficiently strong @xmath243
, the chain assumes a collapsed state , while @xmath245 corresponds to good solvent .
we used a combination of stochastic dynamics @xcite and the pivot algorithm @xcite . applying standard methods @xcite on data of chains of length of up to @xmath246
, we located the @xmath8 point at @xmath247 . in the good solvent limit , and at @xmath248 , we also ran an @xmath249 chain .
_ model b _ is a mesoscopic model for an aqueous solution of the sodium salt of poly ( acrylic acid ) ( paa ) , whose input parameters have been derived from an extensive atomistic simulation of an aqueous paa solution ( @xmath250 k and @xmath251 atm ) in the highly diluted regime , such that the ion concentration ( number of charges on the chain , plus counterions ) is @xmath252 mol / l @xcite . from this simulation
, structural averages like the distributions of bond angles or radial distribution functions between monomers were extracted .
we mapped this system to the mesoscale by replacing one repeating unit ( i. e. one monomer ) by one bead .
as center of the coarse grained ( cg ) beads , the monomer center of mass ( excluding the sodium ion ) was chosen .
bonded as well as non bonded terms were parameterized by systematically varying the interactions until the structure of the atomistic model was reproduced @xcite .
this also allowed us to neglect all explicit water molecules and sodium ions ( necessarily present in the parent atomistic simulation ) in subsequent cg simulations .
their effect on the paa chain conformation is , however , implicitly present in the model .
this means that a system of roughly @xmath253 atoms could be reduced to a system which consists of only @xmath254 `` super atoms '' . as in model
a , we used both stochastic dynamics and pivot monte carlo moves .
the final force field was utilized to calculate @xmath7 , @xmath0 and other static properties like the structure factor for paa strands of length 8 to 3155 repeating units @xcite .
the numerical results agree well with light scattering data on dilute paa solutions with corresponding mean molar weights .
in particular , the hydrodynamic radii of six different paa salt samples with molecular weights in the range from @xmath255 to @xmath256 g / mol were measured . for four samples , the molar masses @xmath257 and
the radii of gyration @xmath7 were measured as well .
the paa samples were of polydispersity @xmath258 between @xmath259 and @xmath260 and diluted in aqueous nacl containing solution ( @xmath261 mol / l ) @xcite . finally , _
model c _ is the saw on the face centered cubic lattice , which we prefer over the simple cubic for reasons of increased local flexibility , which in turn means proximity to the saw fixed point .
units of length are defined in such a way that the bond length is @xmath262 .
the chains of length @xmath263 were generated by using a dimerization procedure @xcite .
up to @xmath264 the statistical sample always consisted of @xmath265 chains , while @xmath266 for @xmath267 and @xmath268 for @xmath269 . in
what follows , we outline our @xmath7 and @xmath0 data for these three models .
figures [ fig : martintheta ] and [ fig : martingood ] summarize our results for model a at @xmath8 condition @xmath248 ( fig .
[ fig : martintheta ] ) and at good solvent condition @xmath245 ( fig .
[ fig : martingood ] ) , respectively . for the @xmath8 chains
, we obtained very good fits with the functions @xmath270 and @xmath271 , while for the good solvent data the analogous fits are @xmath272 and @xmath273 .
these fit functions are also shown in figs .
[ fig : martintheta ] and [ fig : martingood ] .
the ratio @xmath274 , as it results from these data , is shown in the subsequent figs .
[ fig : ratiomartintheta ] and [ fig : ratiomartingood ] for @xmath8 and good solvent conditions , respectively .
it should be noted that the numerical resolution ( for each of our models ) is clearly by far not competitive with the study by li _
_ @xcite . for this reason ,
we did not attempt to determine the exponents from our data , but rather used the values for @xmath87 and @xmath37 from ref . .
we did not include an @xmath54 term in the fit for @xmath0 in the saw case , although such a term is expected to be present .
the reason is that our model a data are too inaccurate to allow for such a three
parameter fit in a stable way .
similarly , we ignored the non analytic corrections to scaling in the @xmath8 case , for essentially the same reason , as has been discussed in some more detail in sec .
[ sec : intro ] .
taking the statistical inaccuracies of the data , and of the resulting fit parameters into account , we obtain for the asymptotic amplitude ratio @xmath274 the values @xmath275 at the @xmath8 point , and @xmath276 in the excluded volume case . the actual error in @xmath82
is expected to be significantly larger , since neither the uncertainties in the exponents and in the location of the @xmath8 point , nor systematic errors due to higher order corrections to scaling have been taken into account .
this is particularly apparent in the @xmath8 case , where one expects in the asymptotic long chain limit rather the gaussian value @xmath10 , but also obvious in the saw case , where the results on the longer chains of model c yield a considerably smaller value for @xmath82 .
the most interesting aspect of model b is that it closely resembles a real system , and a quantitative comparison with experiments is possible @xcite . in fig .
[ fig : rgrh_dirk ] we show simulation results for @xmath7 , @xmath0 , and their ratio .
the data are taken as published in ref . .
for the ratio , experimental results are also included .
the scaling @xmath4 , and the extrapolation to @xmath277 is nicely borne out by the simulation data .
the experiments are too inaccurate to demonstrate a clear systematic trend . in spite of this ,
an extrapolation yields @xmath278 , which means that the theoretical calculations are supported by data of a real chemical system .
our model c data ( saw ) comprise the largest range of chain lengths of our three models , combined with precise estimates of statistical errors , which allows a more detailed data analysis . for our @xmath7 data
, we obtained the fit @xmath279 with @xmath280 , @xmath281 , where we again use the exponents from ref . .
the deviation @xmath282 ( sum of the residuals squares , normalized by the variances ) has the value @xmath283 ( 10 data points ) . the corresponding quality of fit @xmath284 , which is the probability to observe the measured @xmath282 value , or a larger one , is @xmath285 . our data , in a representation which emphasizes the corrections to scaling ,
are shown in fig .
[ fig : bdrg2corscal ] .
it is seen that these are indeed weak , highlighting the difficulties in determining an accurate value for the correction to scaling exponent . turning to our @xmath0 data from model c
, we first did a nonlinear two parameter fit @xmath286 , resulting in @xmath287 .
however , this fit is very poor , with a least
square sum @xmath288 .
conversely , a linear two parameter fit @xmath289 yields a rather good value @xmath290 ( @xmath291 ) , with @xmath292 , @xmath293 , demonstrating also numerically that @xmath0 data should be interpreted in terms of corrections to scaling , instead of an effective exponent .
actually , one should expect the presence of an additional correction of order @xmath54 , @xmath294 .
since this correction tends to decrease @xmath7 ( see fig . [
fig : bdrg2corscal ] ) , it should also decrease @xmath0 , i. e. increase @xmath295 , or weaken the analytic @xmath4 term .
thus , in a regression @xmath296 , where we keep @xmath143 fixed , one should obtain the best fit for a value of @xmath143 slightly smaller than unity .
this is indeed what we observe , as seen from fig .
[ fig : bdqvalues ] , where we plot the quality @xmath284 of such a fit as a function of @xmath143 .
this figure also clearly rules out a single correction to scaling with an exponent of @xmath96 or even larger .
we thus attempted a three
parameter fit @xmath297 to also take the @xmath54 term into account .
the result of this fit , which seems to be reasonably stable , is @xmath298 , @xmath299 , @xmath300 , with @xmath301 , and a very good quality @xmath302 .
we thus use this fit to demonstrate the corrections to scaling of @xmath155 in fig .
[ fig : bdrhcorscal ] , where the presence of the @xmath54 term shows up in a slight curvature . finally , we also used this fit , combined with the corresponding one for @xmath7 ( see fig .
[ fig : bdrg2corscal ] ) , to describe the data on the ratio @xmath274 , as shown in fig .
[ fig : bdratio ] , where the asymptotic value is @xmath303 .
again we feel that the real uncertainty is larger , due to lack of control of the systematic errors .
we also checked that both the quality of fit , and the value of @xmath82 did not change significantly when we reduced the exponent @xmath37 to its theoretical value @xcite @xmath304 . to summarize ,
we have collected our most important numerical results , the extrapolated @xmath82 values , in table [ tab : rgrhvalues ] .
.asymptotic universal ratio @xmath6 as estimated by numerical simulations of various models ( see text ) .
error bars take into account statistical uncertainties only , while systematic errors in the extrapolation procedure are neglected . [
cols="^,^",options="header " , ]
stimulating discussions with j. horbach , a. j. c. ladd , j. j. de pablo , and s. wiegand are gratefully acknowledged .
we thank l. schfer , a. z. akcasu and y. tsunashima for useful remarks and hints to the literature , g. besold for a critical reading of the manuscript , and dsm and the bmbf competence center in materials simulations for financial support .
quite usually , sums are approximated via the corresponding integrals . the euler
maclaurin formula @xcite , which we outline here for the convenience of the reader , constructs a systematic asymptotic expansion around that approximation . defining a difference operator @xmath37 via @xmath305
one obviously has @xmath306 for @xmath307 and thus @xmath308 on the other hand , @xmath309 or @xmath310^{-1 } \nonumber \\ & = & \int dx \sum_{k = 0}^\infty \frac{b_k}{k ! } \left ( \frac{d}{dx } \right)^k , \end{aligned}\ ] ] where @xmath311 are the bernoulli numbers defined via the taylor expansion of @xmath312 : @xmath313 , @xmath314 , @xmath315 , @xmath316 , , @xmath317 .
hence , @xmath318 and thus @xmath319 where the `` integration '' constant is determined via ( perhaps numerical ) comparison of both sides . for a power law with @xmath320
one thus finds from the definition of the riemann zeta function @xmath321 by analytic continuation with respect to @xmath21 , this result holds for general @xmath21 @xcite . | we discuss the hydrodynamic radius @xmath0 of polymer chains in good solvent , and show that the leading order correction to the asymptotic law @xmath1 ( @xmath2 degree of polymerization , @xmath3 ) is an `` analytic '' term of order @xmath4 , which is directly related to the discretization of the chain into a finite number of beads .
this result is further corroborated by exact calculations for gaussian chains , and extensive numerical simulations of different models of good solvent chains , where we find a value of @xmath5 for the asymptotic universal ratio @xmath6 , @xmath7 being the chain s gyration radius . for @xmath8 chains
the data apparently extrapolate to @xmath9 , which is different from the gaussian value @xmath10 , but in accordance with previous simulations .
we also show that the experimentally observed deviations of the initial decay rate in dynamic light scattering from the asymptotic benmouna
akcasu value can partly be understood by similar arguments . 2 |
the appearance of added stability in atomic nuclei at certain nucleon numbers prompted the development of the nuclear shell model .
the robustness of the magic numbers at 2 , 8 , 20 , 28 , 50 , 82 , and 126 has been well demonstrated for nuclei near the valley of @xmath0 stability . however , only limited data are available for @xmath0-unstable nuclei with magic numbers of protons and/or neutrons . in several instances , the observed properties of such radioactive nuclei are not consistent with predictions assuming stabilized proton and/or neutron cores . as examples , the non - existence of @xmath14o@xmath15 @xcite , the appearance of an island of inversion around @xmath16mg@xmath15 @xcite and suspected collectivity at @xmath17ni@xmath18 @xcite all call to question a simple extrapolation of the known magic numbers to the drip - lines of the nuclear chart .
the ca isotopes have a closed - shell number of protons ( @xmath19 ) and have been well studied in the range of the doubly - magic nuclei @xmath20ca@xmath15 and @xmath21ca@xmath18 , where neutrons fill the @xmath22 orbital . beyond doubly - magic @xmath21ca ,
neutrons begin filling the upper @xmath12 shell .
neutron transfer data @xcite are available for levels in @xmath23ca@xmath24 , and the deduced spectroscopic factors signify energy gaps between the adjacent @xmath25 , @xmath26 , and @xmath27 single - particle orbitals .
such gaps suggest added stability for the ca isotopes with @xmath28 and @xmath29 .
indeed , a subshell closure at @xmath28 for @xmath30ca @xcite , @xmath1ti @xcite , and @xmath5cr @xcite has been inferred from the systematic variation of first - excited @xmath31 states in the even - even isotopes of these elements as a function of neutron number .
the yrast structures of the even - even cr @xcite and ti @xcite isotopes at higher spins also show irregular energy spacings at @xmath13 that can be understood if a significant gap in neutron single - particle states between @xmath32 and @xmath33 , @xmath34 exists .
additional support for the apparent shell gap at @xmath13 is found in the small @xmath35 values measured for @xmath1ti@xmath36 @xcite and @xmath5cr@xmath36 @xcite .
no substantial experimental evidence for an expected @xcite subshell closure at @xmath37 has been found to date , as the even - even cr @xcite and ti @xcite isotopes show a decrease in energy of the @xmath38 level when moving from @xmath13 to @xmath37 .
little data are available on the structures of the neutron - rich ca isotopes beyond @xmath30ca@xmath36 .
knowledge of existence has only been extended to @xmath5ca@xmath39 @xcite .
the heaviest ca isotope where structure and half - life information is available is @xmath8ca .
the @xmath0-decay half - life of @xmath8ca was deduced to be @xmath40 ms @xcite . in that study , @xmath8ca was produced from the decay of @xmath8k , which was made directly by fragmentation of an ir target by light ions , and extracted from an ion source .
the mass - separated sample was studied via @xmath0-delayed neutron spectroscopy .
a two - component decay curve was observed , and the component with the longer decay constant was attributed to the daughter @xmath8ca decay . a delayed neutron branch of ( @xmath41)% was obtained for @xmath8ca .
a subsequent @xmath0-decay study of @xmath8k @xcite revealed a single bound state at 2.2 mev above the @xmath8ca ground state .
significant neutron branching was also observed for the decay of @xmath8k , but no additional information on the @xmath0-decay properties of the @xmath8ca daughter was reported in ref .
@xcite .
we report results of @xmath0-delayed @xmath10-ray spectroscopic measurements of @xmath42ca to examine the robustness of the @xmath43 proton shell closure for very neutron - rich ca isotopes .
the half - lives of @xmath44ca were determined for the first time , and a longer half - life value was deduced for @xmath8ca .
the systematic trend in half - life values for the neutron - rich ca isotopes is compared to predictions from gross @xmath0 decay and shell model calculations completed in the full @xmath12-shell model basis .
neutron - rich ca nuclei were produced by fragmentation of a 140 mev / nucleon @xmath45ge beam at national superconducting cyclotron laboratory at michigan state university .
the primary beam was incident on a 352 mg/@xmath46 be target at the object position of the a1900 fragment separator @xcite .
two independent settings of the a1900 separator were used to separate the desired isotopes from unwanted reaction products .
the first setting was optimized for production of @xmath1k with magnetic rigidities b@xmath47 tm and b@xmath48 tm , and included @xmath49ca .
the second a1900 setting with b@xmath50 tm and b@xmath48 tm was chosen to center the production of @xmath1ca at the a1900 focal plane without changing the tune of the second half of the spectrometer ( and the downstream beam transport magnets ) .
the particle identification spectra collected at the experimental end station for both a1900 settings are presented in fig .
the full momentum acceptance of the a1900 ( @xmath51% ) was used for fragment collection .
an al wedge of thickness 300 mg/@xmath46 was placed at the intermediate momentum - dispersive image to provide differential energy loss to fragments of differing @xmath52 traversing the spectrometer .
a thin plastic scintillator detector was also placed near this position , and position information deduced from this detector was used to correct the fragment time of flight .
k and ( b ) @xmath1ca were centered in the dispersive plane at the final focus of the device .
the energy loss measurement is from the most upstream si pin detector of the beta counting system with thickness 991 @xmath53 m . ]
the experimental end station consisted of detectors from the beta counting system @xcite and the segmented germanium array @xcite .
the fragments transported from the a1900 to the experimental end station first encountered a stack of three si pin detectors with thicknesses 991 , 997 , and 309 @xmath53 m .
these detectors provided energy loss information on the incoming beam , critical for particle identification .
the timing signal from the most upstream si pin was used in conjunction with the cyclotron radiofrequency to deduce fragment time of flight information .
the si pin stack was followed by a si double - sided multistrip detector ( dssd ) and six si single - sided multistrip detectors ( sssds ) .
the dssd had thickness 979 @xmath53 m and was segmented into 40 strips on both front and back , for a total of 1600 pixels .
the sssds had nominal thicknesses of 1 mm and were segmented into 16 strips .
the sssds strip orientation alternated between horizontal and vertical ( relative to ground ) for each successive detector to provide two - dimensional position information for particles moving downstream of the dssd .
the majority of fragments were stopped in either the dssd or the most upstream sssd ; a consequence of the large momentum acceptance of the a1900 separator .
sixteen detectors from the segmented germanium array were arranged in two concentric rings around the vacuum beam line surrounding the si detectors of the beta counting system .
the @xmath10-ray peak detection efficiency was 20% at 100 kev and 7% at 1 mev .
the energy resolution of each ge detector in the array was @xmath54 kev for the 1.3 mev @xmath10-ray transition in @xmath55co .
the master trigger for event readout was determined by a logical or of any of the following three conditions : 1 ) signal above threshold in any of the 16 strips of the most upstream sssd ; 2 ) signal above threshold in any of the 40 front strips and any of the 40 back strips of the dssd ; 3 ) signal above threshold in the most upstream si pin detector ( used for prompt @xmath10-ray detection ) .
@xmath10 rays detected within 15 @xmath53s of a master trigger were recorded into the event stream .
the long gating time for @xmath10 rays enabled identification of potential microsecond isomers populated either in the fragmentation process or following @xmath0 decay .
each event was tagged with an absolute time stamp based on a 50 mhz oscillator that was input to a 32-bit free - running scaler .
details of an implantation event , including energy loss and time - of - flight information required for particle identification , were stored in a two - dimensional array with indices based on pixel location .
correlations with the energy - loss signal from @xmath0-decay electrons detected in the dssd were then processed , based on temporal and position information .
position correlations were restricted to @xmath0-decay electrons that emanated from the same pixel or one of the surrounding eight pixels of an implantation .
time correlations had two defined limits : a maximum time for a single @xmath0-decay event for a given implantation event and a minimum time for back - to - back implantation events into the same pixel .
the details of the implantation event were cleared from the two - dimensional array when these defined limits were not satisfied .
the results presented in the next section were obtained with a maximum correlation time of @xmath56 s and a minimum time for back - to - back implantations of @xmath56 s.
the decay curves for @xmath42ca are given in fig .
the half - life of each nuclide was deduced by fitting the decay curves to a function that considered exponential decay of the parent , exponential growth and decay of the daughter , and constant background .
decay information on the daughter nuclide was taken from the literature , with additional details provided in the following subsections .
table [ tab1 ] summarizes the half - life results for @xmath42ca .
ca ; ( b ) @xmath1ca ; ( c ) @xmath3ca ; and ( d ) @xmath5ca .
the fitting function ( solid black line ) considered contributions from parent exponential decay ( dotted line ) , daughter exponential growth and decay ( dashed line ) , and a flat background ( solid gray line ) . the only fitted curve that resulted in a non - negligible value for the background term was that for @xmath5ca .
additional details regarding the fitting procedure used to extract half - life values are provided in the text . ]
ccccc & & + + nuclide&implanted&correlated&this work&previous + @xmath8ca & 209 & 32 & @xmath57 & @xmath40 + @xmath1ca & 654 & 136 & @xmath58 & + @xmath3ca & 246 & 52 & @xmath59 & + @xmath5ca & 99 & 18 & @xmath60 & + .
the spectra of @xmath10 rays correlated with the @xmath0 decays of @xmath42ca are presented in fig .
@xmath1ca is the only nuclide where a delayed @xmath10 ray is assigned to the parent @xmath0 decay .
no evidence for isomeric @xmath10-ray emission within 15 @xmath53s of an implantation was found for the neutron - rich @xmath42ca isotopes .
-ray spectrum in the range 50 to 1600 kev correlated with @xmath0-decay events for : ( a ) @xmath8ca ; ( b ) @xmath1ca ; ( c ) @xmath3ca ; and ( d ) @xmath5ca .
observed transitions are marked by their energy in kev . ]
more detailed discussions are included below on the results for each isotope .
the half - life of @xmath8ca reported here , @xmath61 ms , is significantly longer than the value @xmath40 ms reported in ref .
the previous half - life value was deduced following the decay chain @xmath8k @xmath9 @xmath8ca @xmath9 @xmath8sc , where @xmath8k was directly produced by proton spallation of an ir target and only @xmath0 particles in coincidence with delayed neutrons were recorded . in that work ,
data collected in the range 300 to 700 ms were fitted with a single exponential decay , assuming contributions from neither the parent nor the grand - daughter over this time range .
these assumptions are reasonable considering the short half - life of @xmath8k , @xmath62 ms @xcite , and the unknown , but expected long ( @xmath63 s ) half - life of @xmath8sc @xcite .
the decay curve presented in fig . [ fig2 ] considers only direct production of @xmath8ca , with no ambiguity regarding the fit to a convoluted spectrum involving both parent and daughter decays . in the present work , the decay curve derived from @xmath0-delayed neutron events , presented in fig . 2 of ref .
@xcite , was refitted with a function based on the batemann equations , where the parent exponential decay and daughter exponential growth and decay were considered .
the parent half - life value was fixed at @xmath64 ms , and two fixed values of the half - life of the @xmath8ca were used : the @xmath65-ms value deduced by langevin _
_ @xcite and the value of @xmath66 ms deduced here .
the parent activity was taken as a free parameter , and the daughter activity was determined as 40% of the parent activity , based on the measured delayed - neutron branching following @xmath0 decay of this nuclide @xcite .
the resulting fits are compared with the data in fig .
[ fig4 ] . an @xmath67 regression analysis to check the goodness of fit produced values of 0.952 and 0.924 for the functions that used @xmath8ca half - life values of @xmath65 ms and @xmath66 ms , respectively .
the difference in @xmath67 values between the two fits is not significant , if the errors in the half - life values of the parent ( 17% ) and daughter ( @xmath68 20% ) are considered .
the longer half - life value deduced here for @xmath8ca appears consistent with the @xmath0-delayed neutron decay data obtained in the previous study of the @xmath69 decay chain .
-delayed neutron decay after collection of @xmath8k taken from ref .
data were fitted with two different half - life values of the daughter @xmath8ca .
regression analysis results in @xmath67 values of 0.952 and 0.924 for the fits with @xmath8ca half - life values of 90 ms and 230 ms , respectively . ]
however , it should be acknowledged that we can not rule out that the longer half - life measured here could result from the presence of a second @xmath0-decaying state in @xmath8ca .
isomeric @xmath70 states are known in the neutron - rich @xmath71fe@xmath72 @xcite and @xmath73cr@xmath72 @xcite isotopes .
both the magnetic dipole moment @xcite and electric quadrupole moment @xcite suggest the @xmath74 level in @xmath71fe is characterized by a deformed potential .
although the neutron - rich ti and ca isotopes should show less deformed structures due to their proximity to the @xmath19 shell closure , the possible influence of the @xmath75 orbital on the low - energy structure of these isotopes requires further investigation .
no previous information on the half - life of @xmath1ca was available .
the deduced half - life , @xmath76 ms , was obtained by fitting the fragment-@xmath0 decay curve in fig .
[ fig2 ] to a function that included exponential decay of the parent and exponential growth and decay of the @xmath1sc daughter ( the constant background term was small and could be neglected ) .
the @xmath1sc half - life value was fixed at @xmath77 ms @xcite , while the parent half - life and activity values were free parameters . a single @xmath10 ray with energy @xmath78 kev has been assigned to the decay of @xmath1ca based on the delayed @xmath10-ray spectrum presented in fig . [
fig3](b ) .
the decay curve constructed from @xmath1ca-@xmath0 correlations with an additional requirement of a coincidence with a 247-kev @xmath10 ray is given in fig .
the fitted decay curve returns a half - life consistent with that deduced from the fragment-@xmath0 correlation data .
ca showing fragment-@xmath0 correlations with an additional requirement of a 247-kev @xmath10 ray in coincidence .
the data were fitted with a single exponential decay with fixed parent half - life ( @xmath79 ms ) and the initial activity as a free parameter . ]
the absolute intensity of the 247-kev transition was determined to be @xmath80% , based on a gaussian fit to the @xmath10-ray peak in fig .
[ fig3](b ) , the measured peak efficiency curve for the 16-detector @xmath10-ray array , and the number of @xmath1ca implantations correlated with @xmath0 decay ( see table [ tab1 ] ) .
the absolute intensity implies that a majority of the apparent @xmath0 intensity from the decay of @xmath1ca proceeds through an excited state depopulated by the 247-kev transition .
this transition is proposed to directly feed the ground state of @xmath1sc .
the new 247-kev excited state is tentatively assigned spin and parity @xmath81 , based on selection rules for allowed @xmath0 decay .
the uncertainty in the spin and parity assignment for the 247-kev level arises from the large @xmath82 value associated with the @xmath1ca @xmath0 decay .
the low statistics in the delayed @xmath10-ray spectrum in fig .
[ fig3](b ) can not exclude the presence of higher energy , lower intensity transitions that may directly feed the level at 247 kev .
the @xmath0-decay sequence and known levels in @xmath1sc are displayed in fig .
the decay @xmath82-value is taken from ref .
. the excited state at 110 kev in @xmath1sc was identified as a @xmath53s isomer in studies by grzywacz _
_ @xcite and was reaffirmed in ref .
the spin and parity of the @xmath1sc ground state was limited to @xmath83 , based on the @xmath0-decay feeding pattern to excited states in @xmath1ti @xcite .
placement of the 247-kev transition would preclude @xmath84 for the final state of this``prompt '' @xmath10-ray transition , therefore , spin and parity @xmath85 is proposed for the ground state of @xmath1sc .
the 247-kev transition is presumed to have @xmath86 multipolarity , and the estimated half - life from weisskopf estimates of 51 ns is consistent with the `` prompt '' nature of this transition .
sc and @xmath0-decay scheme for the parent nuclide , @xmath1ca .
the energies of each state are given in kev .
the level with energy 110 kev was previously identified as a microsecond isomer by grzywacz _
_ @xcite and later confirmed by liddick _
the 110-kev @xmath10-ray was not observed in the @xmath0-decay study of @xmath1ca presented here . ]
the only previous information regarding the isotopes @xmath49ca was their existence , established from fission of relativistic @xmath87u ions @xcite . in that study , six and three counts were associated with the production of @xmath3ca and @xmath5ca , respectively . we have achieved sufficient statistics in the present study to deduce half - life values for @xmath49ca for the first time .
the fragment-@xmath0 decay curves were fitted to a function that included the exponential decay of the parent , the growth and decay of the daughter , and a linear background . for the @xmath3ca decay curve in fig .
[ fig2](c ) , the @xmath3sc daughter half - life was fixed to a value @xmath88 ms .
this new half - life was obtained in the present work from the analysis of @xmath0-decay events correlated with @xmath3sc implantations shown in fig .
[ fig7 ] , and represents a significant improvement in precision with respect to previously reported values @xcite .
the parent @xmath3ca half - life , initial parent activity , and constant value for background were taken as free parameters .
the grand - daughter @xmath3ti has a known half - life of @xmath89 s @xcite and was not considered in the curve fitting function .
the fit resulted in a half - life of @xmath90 ms for @xmath3ca ( see fig .
[ fig2 ] ) , with a negligible background contribution .
particles correlated with @xmath3sc implantation events .
the decay data were fitted to a function ( solid black line ) that considered the exponential decay of the @xmath3sc parent ( dotted line ) and exponential growth and decay of the @xmath3ti daughter ( dashed line ) with known half - life @xmath89 s @xcite . ] a single @xmath10-ray transition of energy 592 kev was evident in the @xmath0-delayed @xmath10-ray spectrum correlated with @xmath3ca implantations shown in fig .
this transition was previously assigned to the decay of the @xmath3sc daughter @xcite .
the fit to the @xmath5ca decay curve in fig .
[ fig2](d ) considered a fixed value of @xmath91 ms for the half - life of the @xmath5sc daughter .
there are , in fact , two @xmath0-decaying states in @xmath5sc @xcite .
decay of the @xmath5ca ground state , with presumed spin and parity @xmath92 , would populate lower - spin states in the daughter nucleus .
therefore , the dominant decay constant to consider for fitting the @xmath5ca decay curve is that associated with the low - spin ( @xmath81 ) level in @xmath5sc . the half - life deduced for @xmath5ca @xmath0 decay was @xmath93 ms , about a factor of two shorter than that in the neighbor @xmath3ca .
no evidence was found for delayed @xmath10-ray transitions correlated with @xmath5ca @xmath0 decays
. however , the small number of correlated decays ( 18 total decay events see table [ tab1 ] ) combined with the measured @xmath10-ray efficiency curve would preclude observation of more than one or two counts for @xmath10 rays with energies at the maximum of the ge array efficiency curve .
the @xmath0 decay of @xmath1ca apparently proceeds to a single state in @xmath1sc residing 247 kev above the ground state .
this state has tentatively been assigned @xmath81 spin and parity .
this new information for @xmath1sc can be combined with previous knowledge on isomeric and @xmath0 decay to better understand the coupling of valence protons and neutrons in this odd - odd nucleus . the known low - energy levels in @xmath1sc in fig . [ fig8 ]
are compared with results of shell model calculations completed in the full @xmath12-shell model basis with the interactions gxpf1 @xcite , gxpf1a @xcite , and kb3 g @xcite .
the difference in calculated level orderings between the gxpf1 and kb3 g interactions can be related to the relative energies of the neutron single - particle @xmath94 and @xmath95 orbitals .
the @xmath85 , @xmath96 doublet near the ground state in @xmath1sc in the calculations with the gxpf1 and gxpf1a interactions arises from the coupling of the @xmath33 and @xmath97 orbitals .
this configuration appears lowest in energy in @xmath1sc since the effective single - particle energy of the @xmath33 orbital is below that of the @xmath34 orbital with the gxpf1 and gxpf1a interactions .
the first @xmath81 level is attributed mainly to the coupling of the @xmath34 and @xmath97 single - particle orbitals .
this level is at the ground state in the shell - model results with the kb3 g interaction , as this interaction places the effective single - particle energy of the @xmath33 orbital above that of the @xmath34 orbital .
the position of the tentative @xmath81 level at 247 kev assigned in this work , relative to the @xmath98 ground state , is not well reproduced by either the kb3 g , gxpf1 , or gxpf1a interactions .
sc compared to results of shell model calculations employing the gxpf1 @xcite , gxpf1a @xcite , and kb3 g @xcite effective interactions . ]
the gxpf1 interaction proved successful in reproducing the yrast structures of the ti isotopes up to @xmath5ti@xmath99 , where the first excited @xmath31 state was found to have an energy of 1129 kev @xcite , nearly 400 kev below the gxpf1 expectation .
modification of three two - body pairing matrix elements associated with the @xmath94 , @xmath100 , and @xmath95 orbitals and two quadrupole - quadrupole interaction matrix elements coupling the @xmath94 and @xmath95 orbitals to @xmath101 and @xmath102 resulted in the gxpf1a interaction @xcite .
the shell model results with the gxpf1a interaction reproduced the low - energy yrast structure of @xmath5ti @xcite better , while maintaining good agreement with the known level structures of the less neutron - rich ti isotopes .
as noted above , neither interaction does well in reproducing the position of the tentative @xmath81 state in @xmath1sc at 247 kev , which should be dominated by the shell model configuration @xmath103 .
the low energy of the @xmath104 level relative to the proposed @xmath98 ground state , with dominant configuration @xmath103 , suggests that while the level ordering of the @xmath95 and @xmath94 neutron single - particle orbitals in the gxpf1 and gxpf1a interactions is likely correct , the effective energy spacing between the neutron @xmath95 and @xmath94 orbitals is overestimated or the interaction strength between @xmath105 and @xmath106 is underestimated .
another outstanding question with regards to the low - energy structure of @xmath1sc is the tentative spin - parity assignment for the isomeric state located 110 kev above the ground state .
@xcite deduced @xmath86 multipolarity for the 110 kev @xmath10 ray , based on a comparison of the experimental half - life value @xmath107s with single - particle estimates .
tentative spin and parity @xmath108 was then assigned to the 110-kev excited state . in ref .
@xcite , it was suggested that the @xmath86 isomer may represent the decay of the first excited @xmath81 state , whose energy may be lower than predicted by the shell model calculations utilizing the gxpf1 interaction .
indeed , the @xmath109 state is lower than predicted , tentatively established in @xmath1sc at 247 kev .
regardless of the interaction employed , the first excited @xmath108 state , associated with the @xmath103 configuration , is expected at an energy in excess of 1.5 mev ( see fig . [ fig8 ] ) .
the shell model results all suggest a doublet of levels with @xmath110 near the @xmath1sc ground state .
these two levels belong to the odd - odd @xmath111 multiplet .
the transition probabilities @xmath112 and @xmath113 calculated with the gxpf1 and gxpf1a interactions are listed in table [ tab2 ] , along with the calculated half - lives .
the @xmath114 strength dominates , but the calculated partial half - life was in the range of a few nanoseconds , much longer than the 9.3 ps half - life calculated from single - particle estimates .
the @xmath114 hindrance factor appears significant for this @xmath10-ray transition .
therefore , the long ( @xmath53s ) half - life established for the isomeric transition in @xmath1sc does not exclude possible @xmath96 spin and parity for the first - excited state .
.transition probabilities and partial half - lives for the @xmath10-ray transition between the @xmath85 and @xmath96 members of the @xmath111 multiplet in @xmath1sc .
the experimental transition energy of 110 kev was used in the calculations . [ cols="^,^,^,^,^,^,^ " , ] .
the @xmath0 decay of @xmath8ca from the presumed @xmath115 ground state is predicted to feed the first excited @xmath116 state in @xmath8sc with intensity 60% of the total feeding strength .
the @xmath0 strength to neutron unbound states is calculated to be 6% , which is smaller than that observed in ref .
. the expected decay of @xmath1ca is similar to that of @xmath8ca . about 60% of the calculated @xmath0-decay strength
is predicted to feed the first @xmath81 state in @xmath1sc , and the @xmath117 value is also calculated to be 6% . the observed @xmath0-decay strength from @xmath1ca
is apparently dominated by decay to the the 247-kev level in @xmath1sc .
the absolute intensity of the 247-kev @xmath10-ray transition that depopulates this state is 97% , with no other peak ( intensity threshold @xmath118 below 1 mev ) in the delayed @xmath10-ray spectrum displayed in fig .
[ fig3](b ) .
however , as noted earlier , only tentative @xmath81 spin and parity have been proposed for this 247-kev level , since the low statistics of the measurement does not rule out the presence of higher - energy @xmath10 rays feeding the 247-kev level from potential higher - lying @xmath81 states .
the shell model results with the gxpf1 interaction support the notion that the primary branching in the @xmath0 decay of @xmath1ca is to the first excited @xmath81 in @xmath1sc .
the experimental delayed - neutron branching reported in table [ tab3 ] for @xmath1ca is governed by the lower limit of the absolute intensity of the 247-kev @xmath10 ray .
the @xmath0-decay properties of @xmath49ca were found to be similar to the shell model results , as was the case with @xmath119ca .
the decays are dominated by fast ground state to ground state transitions , which carry more than 90% of the @xmath0-decay strength . since only a small number of decays were detected for both @xmath3ca and @xmath5ca , there was little possibility to observe branching to excited states in the @xmath3sc and @xmath5sc daughters , respectively , that would lead to @xmath10-ray or neutron emission .
the short half - life values and large , direct ground - state @xmath0 branching again reflect the dominance of the spin - flip @xmath120 decay process in the shell model results . the good agreement between the experimental and shell model half - lives for @xmath49ca support the notion that the @xmath34 orbital is near the fermi surface for these nuclides with @xmath121 .
the @xmath0-decay properties of the neutron - rich ca isotopes have been extended to @xmath5ca@xmath39 . the half - lives of @xmath44ca have been measured for the first time to be @xmath76 ms , @xmath90 ms , and @xmath93 ms , respectively . a new half - life of @xmath57 ms has been deduced for @xmath8ca , a value longer than that reported previously .
the trend of the @xmath0-decay half - lives of the neutron - rich ca isotopes shows a regular decrease except at @xmath30ca@xmath36 , which is attributed to the known subshell closure at @xmath13 .
the low - energy structure of @xmath1sc , the @xmath0-decay daughter of @xmath1ca , has been elucidated further .
the ground and first excited states most likely arise from coupling of the @xmath97 and @xmath33 orbitals , suggesting that the @xmath33 single - particle orbital resides below the @xmath34 state at @xmath122 .
such level ordering was also inferred from the low - energy structure of the @xmath122 isotone @xmath3ti @xcite . the first @xmath81 level in @xmath1sc has been proposed at an energy of 247 kev above the ground state .
this tentative @xmath81 level is presumed to be a member of the @xmath123 multiplet , and is nearly 800 kev lower than shell model predictions with the gxpf1 and gxpf1a effective interactions .
it seems that the effective energy gap between adjacent neutron single - particle orbitals @xmath94 and @xmath95 is overestimated or the interaction strength between the @xmath97 and @xmath34 orbitals is underestimated by the gxpf1 interaction .
the fast @xmath0 decays of @xmath49ca agree within a factor of two with the shell model results , where the @xmath0 strength is dominated by the @xmath120 spin - flip decay .
the shell model prediction for the half - life of @xmath1ca is also within a factor of two of experiment .
such good agreement may initially invoke support for a shell closure at @xmath37 , since the gxpf1 interaction includes a significant energy gap between adjacent single - particle orbitals @xmath34 and @xmath33 .
however , the @xmath1ca half - life is also found to agree well with the predictions of gross @xmath0 decay theory and the results of qrpa calculations , which do not formally consider a strong shell gap at @xmath37 .
therefore , any interpretation regarding magicity at @xmath1ca@xmath99 will require further experimental investigation .
exploring the @xmath0-decay properties of heavier ca isotopes towards @xmath55ca will prove challenging due to low production cross sections .
however , such studies are vital to better understand the evolution of neutron single - particle states with an increase in isospin and corresponding decrease in neutron separation energy .
the authors thank the nscl operations staff for providing the primary and secondary beams for this experiment and members of the nscl gamma group for assistance in setting up and maintaining the ge detectors .
the work was supported in part by the national science foundation grants phy-06 - 06007 ( nscl ) , phy-02 - 44453 ( jina ) , phy-04 - 56463 ( fsu ) , and phy-05 - 55366 ( cmu / msu ) and the u.s .
department of energy , office of nuclear physics , under contracts de - ac02 - 06ch11357 ( anl ) and de - fg02 - 94-er40834 ( maryland ) .
travel support was also provided by the polish science committee grant p03b 059 29 .
o. tarasov , r. allatt , j.c .
angelique , r. anne , c. borcea , z. dlouhy , c. donzaud , s. grevy , d. guillemaud - mueller , m. lewitowicz , s. lukyanov , a.c .
mueller , f. nowacki , yu .
oganessian , n.a .
orr , a.n .
ostrowski , r.d .
page , yu . penionzhkevich , f. pougheon , a. reed , m.g .
saint - laurent , w. schwab , e. sokol , o. sorlin , w. trinder , and j.s .
winfield , phys .
lett . * 409b * , 64 ( 1997 ) .
t. motobayashi , y. ikeda , y. ando , k. ieki , m. inoue , n. iwasa , t. kikuchi , m. kurokawa , s. moriya , s. ogawa , h. murakami , s. shimoura , y. yanagisawa , t. nakamura , y. watanabe , m. ishihara , t. teranishi , h. okuno , and r.f .
casten , phys .
b * 346 * , 9 ( 1995 ) .
yurkewicz , d. bazin , b.a .
brown , c.m .
campbell , j.a .
church , d.c .
dinca , a. gade , t. glasmacher , m. honma , t. mizusaki , w.f .
mueller , h. olliver , t. otsuka , l.a .
riley , and j.r .
terry , phys .
c * 70 * , 054319 ( 2004 ) .
canada , c. ellegaard , and p.d .
barnes , phys .
c * 4 * , 471 ( 1971 ) .
n. taccetti , p.a .
mando , p. sona , and k.p .
lieb , j. phys .
g * 13 * , 393 ( 1978 ) .
y. uozumi , o. iwamota , s. widodo , a. nohtomi , t. sakae , m. matoba , m. nakano , t. maki , and n. koori , nucl .
* a576 * , 123 ( 1994 ) .
a. huck , g. klotz , a. knipper , c. miehe , c. richard - serre , g. walter , a. poves , h.l .
ravn , and g. marguier , phys .
c * 31 * , 2226 ( 1985 ) .
a. gade , r.v.f .
janssens , d. bazin , r. broda , b.a .
brown , c.m .
campbell , m.p .
carpenter , j.m .
cook , a.n .
deacon , d .-
dinca , b. fornal , s.j .
freeman , t. glasmacher , p.g .
hansen , b.p .
kay , p.f .
mantica , w.f .
mueller , j.r .
terry , j.a .
tostevin , s. zhu , phys .
c * 74 * , 021302 ( 2006 ) .
janssens , b. fornal , p.f .
mantica , b.a .
brown , r. broda , p. bhattacharyya , m.p .
carpenter , m. cinausero , p.j .
daly , a.d .
davies , t. glasmacher , z.w .
grabowski , d.e .
groh , m. honma , f.g .
kondev , w. krolas , t. lauritsen , s.n .
liddick , s. lunardi , n. marginean , t. mizusaki , d.j .
morrissey , a.c .
morton , w.f .
mueller , t. otsuka , t. pawlat , d. seweryniak , h. schatz , a. stolz , s.l .
tabor , c.a .
ur , g. viesti , i. wiedenhoever , and j. wrzesinski , phys .
b * 546 * , 55 ( 2002 ) .
prisciandaro , p.f .
mantica , b.a .
brown , d.w .
anthony , m.w .
cooper , a. garcia , d.e .
groh , a. komives , w. kumarasiri , p.a .
lofy , a.m. oros - peusquens , s.l .
tabor , and m. wiedeking , phys .
b * 510 * , 17 ( 2001 ) . d.e .
appelbe , c.j .
barton , m.h .
muikku , j. simpson , d.d .
warner , c.w .
beausang , m.a .
caprio , j.r .
cooper , j.r .
novak , n.v .
zamfir , r.a.e .
austin , j.a .
cameron , c. malcolmson , j.c .
waddington , and f.r .
xu , phys .
c * 67 * , 034309 ( 2003 ) .
s. zhu , a.n .
deacon , s.j .
freeman , r.v.f .
janssens , b. fornal , m. honma , f.r .
xu , r. broda , i.r .
calderin , m.p . carpenter ,
p. chowdhury , f.g .
kondev , w. krolas , t. lauritsen , s.n .
liddick , c.j .
lister , p.f .
mantica , t. pawlat , d. seweryniak , j.f .
smith , s.l .
tabor , b.e .
tomlin , b.j .
varley , and j. wrzesinski , phys .
c * 74 * , 064315 ( 2006 ) .
b. fornal , s. zhu , r.v.f .
janssens , m. honma , r. broda , p.f .
mantica , b.a .
brown , m.p .
carpenter , p.j .
daly , s.j .
freeman , z.w .
grabowski , n.j .
hammond , f.g .
kondev , w. krolas , t. lauritsen , s.n .
liddick , c.j .
lister , e.f .
moore , t. otsuka , t. pawlat , d. seweryniak , b.e .
tomlin , and j. wrzesinski , phys .
c * 70 * , 064304 ( 2004 ) .
dinca , r.v.f .
janssens , a. gade , d. bazin , r. broda , b.a .
brown , c.m .
campbell , m.p .
carpenter , p. chowdhury , j.m .
cook , a.n .
deacon , b. fornal , s.j .
freeman , t. glasmacher , m. honma , f.g .
kondev , j .- l .
lecouey , s.n .
liddick , p.f .
mantica , w.f .
mueller , h. olliver , t. otsuka , j.r .
terry , b.a . tomlin and k. yoneda ,
c * 71 * , 041302 ( 2005 ) a. brger , t.r .
saito , h. grawe , h. hbel , p. reiter , j. gerl , m. grska , h.j .
wollersheim , a. al - khatib , a. banu , t. beck , f. becker , p. bednarczyk , g. benzoni , a. bracco , s. brambilla , p. bringel , f. camera , e. clment , p.
doornenbal , h. geissel , a. grgen , j. grebosz , g. hammond , m. hellstrm , m. honma , m. kavatsyuk , o. kavatsyuk , m. kmiecik , i. kojouharov , w. korten , n. kurz , r. lozeva , a. maj , s. mandal , b. million , s. muralithar , a. neusser , f. nowacki , t. otsuka , zs .
podolyk , n. saito , a.k .
singh , h. weick , c. wheldon , o. wieland , m. winkler , and rising collaboration , phys .
b * 622 * , 29 ( 2005 ) .
m. honma , t. otsuka , b.a .
brown , and t. mizusaki , phys . rev .
c * 65 * , 061301 ( 2002 ) .
o. sorlin , c. donzaud , f. nowacki , j.c .
angelique , f. azaiez , c. bourgeois , v. chiste , z. dlouhy , s. grevy , d. guillemaud - mueller , f. ibrahim , k .-
kratz , m. lewitowicz , s.m .
lukyanov , j. mrazek , yu .- e .
penionzhkevich , f. de oliveira santos , b. pfeiffer , f. pougheon , a. poves , m.g .
saint - laurent and m. stanoiu , eur .
phys . j. a * 16 * , 55 ( 2003 ) . s.n .
liddick , p.f .
mantica , r.v.f .
janssens , r. broda , b.a .
brown , m.p .
carpenter , b. fornal , m. honma , t. mizusaki , a.c .
morton , w.f .
mueller , t. otsuka , j. pavan , a. stolz , s.l .
tabor , b.e .
tomlin , and m. wiedeking , phys .
* 92 * , 072502 ( 2004 ) .
m. bernas , c. engelmann , p. armbruster , s. czajkowski , f. ameil , c. bockstiegel , ph .
dessagne , c. donzaud , h. geissel , a. heinz , z. janas , c. kozhuharov , ch .
miehe , g. munzenberg , m. pfutzner , w. schwab , c. stephan , k. summerer , l. tassan - got , and b. voss , phys .
415b * , 111 ( 1997 ) . m. langevin , c. detraz , d. guillemaud - mueller , a.c .
mueller , c. thibault , f. touchard , g. klotz , c. miehe , g. walter , m. epherre , and c. richard - serre , phys .
b * 130 * , 251 ( 1983 ) .
f. perrot , f. marechal , c. jollet , ph .
dessagne , j .- c .
angelique , g. ban , p. baumann , f. benrachi , u. bergmann , c. borcea , a. buta , j. cederkall , s. courtin , j .-
daugas , l.m .
fraile , s. grevy , a. jokinen , f.r .
lecolley , e. lienard , g. le scornet , v. meot , ch .
miehe , f. negoita , n.a .
orr , s. pietri , e. poirier , m. ramdhane , o. roig , i. stefan , and w. wang , phys.rev .
c 74 , 014313 ( 2006 ) . d.j .
morrissey , b.m .
sherrill , m. steiner , a. stolz , and i. wiedenhver , nucl .
instrum .
methods phys .
b * 204 * , 90 ( 2003 ) .
prisciandaro , a.c .
morton , and p.f .
mantica , nucl .
instrum .
methods phys .
res . a * 505 * , 140 ( 2003 ) .
mueller , j.a .
church , t. glasmacher , d. gutknecht , g. hackman , p.g .
hansen , z. hu , k.l .
miller , p. quirin , nucl .
instrum .
methods phys .
a * 466 * , 492 ( 2001 ) .
o. sorlin , v. borrel , s. grevy , d. guillemaud - mueller , a.c .
mueller , f. pougheon , w. bohmer , k .-
kratz , t. mehren , p. mller , b. pfeiffer , t. rauscher , m.g .
saint - laurent , r. anne , m. lewitowicz , a. ostrowski , t. drfler , and w .- d .
schmidt - ott , nucl .
a632 * , 205 ( 1998 ) .
r. grzywacz , r. beraud , c. borcea , a. emsallem , m. glogowski , h. grawe , d. guillemaud - mueller , m. hjorth - jensen , m. houry , m. lewitowicz , a.c .
mueller , a. nowak , a. plochocki , m. pftzner , k. rykaczewski , m.g .
saint - laurent , j.e .
sauvestre , m. schaefer , o. sorlin , j. szerypo , w. trinder , s. viteritti , and j. winfield , phys .
lett . * 81 * , 766 ( 1998 ) .
freeman , r.v.f .
janssens , b.a .
brown , m.p .
carpenter , s.m .
fischer , n.j .
hammond , m. honma , t. lauritsen , c.j .
lister , t.l .
khoo , g. mukherjee , d. seweryniak , j.f .
smith , b.j .
varley , m. whitehead , and s. zhu , phys .
c * 69 * , 064301 ( 2004 ) .
i. matea , g. georgiev , j.m .
daugas , m. hass , g. neyens , r. astabatyan , l.t . baby , d.l .
balabanski , g. belier , d. borremans , g. goldring , h. goutte , p. himpe , m. lewitowicz , s. lukyanov , v. meot , f. de oliveira santos , yu.e .
penionzhkevich , o. roig , and m. sawicka , phys .
lett . * 93 * , 142503 ( 2004 ) .
n. vermeulen , s.k .
chamoli , j.m .
daugas , m. hass , d.l .
balabanski , j.p .
delaroche , f. de oliveira - santos , g. georgiev , m. girod , g. goldring , h. goutte , s. grevy , i. matea , p. morel , b.s .
nara singh , yu .- e .
penionzkevich , l. perrot , o. perru , s. peru , o. roig , f. sarazin , g.s .
simpson , yu . sobolev , i. stefan , c. stodel , d.t .
yordanov , and g. neyens , phys .
c * 75 * , 051302 ( 2007 ) .
liddick , p.f .
mantica , r. broda , b.a .
brown , m.p .
carpenter , a.d .
davies , b. fornal , t. glasmacher , d.e .
groh , m. honma , m. horoi , r.v.f .
janssens , t. mizusaki , d.j .
morrissey , a.c .
morton , w.f .
mueller , t. otsuka , j. pavan , h. schatz , a. stolz , s.l .
tabor , b.e .
tomlin , and m. wiedeking , phys .
c * 70 * , 064303 ( 2004 ) .
g. audi , a.h .
wapstra , c. thibault , j. blachot , and o. bersillon , nucl .
* a729 * , 1 ( 2003 ) .
mantica , b.a .
brown , a.d .
davies , t. glasmacher , d.e .
groh , m. horoi , s.n .
liddick , d.j .
morrissey , a.c .
morton , w.f .
mueller , h. schatz , a. stolz and s.l .
tabor , phys .
c * 68 * , 044311 ( 2003 ) .
s. zhu , r.v.f .
janssens , b. fornal , s.j .
freeman , m. honma , r. broda , m.p .
carpenter , a.n .
deacon , b.p .
kay , f.g .
kondev , w. krolas , j. kozemczak , a. larabee , t. lauritsen , s.n .
liddick , c.j .
lister , p.f .
mantica , t. otsuka , t. pawlat , a. robinson , d. seweryniak , j.f .
smith , d. steppenbeck , b.e .
tomlin , j. wrzesinski , and x. wangphys .
b * 650 * , 135 ( 2007 ) | @xmath0-decay properties of neutron - rich ca isotopes have been obtained .
half - life values were determined for the first time for @xmath1ca ( @xmath2 ms ) , @xmath3ca ( @xmath4 ms ) , and @xmath5ca ( @xmath6 ms ) . the half - life of @xmath7 ms deduced for @xmath8ca is significantly longer than reported previously , where the decay chain @xmath8k @xmath9 @xmath8ca @xmath9 @xmath8sc was considered . a delayed @xmath10 ray with energy 247 kev was identified following @xmath0 decay of @xmath1ca , and is proposed to depopulate the @xmath11 level in @xmath1sc .
the @xmath0-decay properties compare favorably with the results of shell model calculations completed in the full @xmath12-space with the gxpf1 interaction .
the half - lives of the neutron - rich ca isotopes are also compared with gross @xmath0-decay theory .
the systematic trend of the neutron - rich ca half - lives is consistent with the presence of a subshell gap at @xmath13 . |
the density dependence of nuclear symmetry energy especially at supra - saturation densities is among the most uncertain properties of neutron - rich nuclear matter@xcite .
however , it is very important for nuclear structure@xcite , heavy - ion reactions@xcite and many phenomena / processes in astrophysics and cosmology @xcite . heavy - ion reactions especially those induced by radioactive beams
provide a unique opportunity to constrain the symmetry energy at supra - saturation densities in terrestrial laboratories .
various probes using heavy - ion reactions have been proposed in the literature , see , e.g. , ref.@xcite for the most recent review .
it is particularly interesting to mention that , besides many significant results about the symmetry energy at sub - saturation densities , see , e.g. , refs.@xcite , circumstantial evidence for a rather soft symmetry energy at supra - saturation densities has been reported very recently@xcite based on the ibuu04 transport model@xcite analysis of the @xmath3 data taken by the fopi collaboration at sis / gsi@xcite .
however , many interesting issues remain to be resolved . thus , to constrain tightly and reliably the nuclear symmetry energy especially at supra - saturation densities , much more efforts by both the nuclear physics and the astrophysics communities are still needed . in the present work
, we have the following two main purposes .
firstly , there is an urgent need to verify the conclusion about the soft symmetry energy at supra - saturation densities required to reproduce the fopi @xmath3 data within transport model analyses@xcite . it is better if this test can be done with not only more @xmath3 data but also other sensitive observables in the most neutron - rich reactions possible .
we will thus make predictions using the same transport model@xcite for doing this test .
secondly , it was predicted that the neutron - proton differential flow is another sensitive probe of the high - density behavior of the nuclear symmetry energy@xcite .
however , it is difficult to measure observables involving neutrons .
one question often asked by some experimentalists is whether the triton-@xmath0he pair may carry the same information as the neutron - proton one .
we will try to answer this question quantitatively by coupling the ibuu04 calculations to a phase - space coalescence after - burner .
indeed , we found that , similar to the neutron - proton pair , the triton-@xmath4he relative and differential transverse flows are sensitive to the high - density behavior of the nuclear symmetry energy . they can be used to test indications about the high - density behavior of the symmetry energy observed earlier from analyzing the @xmath3 data .
our study is carried out based on the ibuu04 version of an isospin and momentum dependent transport model and the simplest phase - space coalescence after - burner . for completeness and consistency
we recall here a few major features of the ibuu04 transport model most relevant to the present study . more details of the model can be found in refs .
the single nucleon potential is one of the most important inputs to buu - like transport models for nuclear reactions . in the ibuu04 transport model
, we use a single nucleon potential derived within the hartree - fock approach using a modified gogny effective interaction ( mdi ) @xcite , i.e. , @xmath5 here @xmath6 is the isospin asymmetry of the nuclear medium . in the above @xmath7 ( @xmath8 ) for neutrons ( protons ) and @xmath9 ; @xmath10 ;
@xmath11 is the phase space distribution function at coordinate @xmath12 and momentum @xmath13 .
the parameter @xmath14 was introduced to mimic predictions on the density dependence of symmetry energy @xmath15 by microscopic and/or phenomenological many - body theories .
the parameters @xmath16 and @xmath17 depend on the @xmath14 parameter according to @xmath18 the coefficients in @xmath16 and @xmath17 and the parameters @xmath19 and @xmath20 were obtained by fitting the momentum - dependence of the @xmath21 to that predicted by the gogny hartree - fock and/or the brueckner - hartree - fock ( bhf ) calculations , the saturation properties of symmetric nuclear matter and the symmetry energy of about @xmath22 mev at normal nuclear matter density @xmath23 @xmath24 .
the incompressibility @xmath25 of symmetric nuclear matter at @xmath26 is set to be @xmath27 mev consistent with the latest conclusion from studying giant resonances @xcite .
the last two terms in eq .
( [ mdi ] ) contain the momentum - dependence of the single - particle potential .
the momentum dependence of the symmetry potential stems from the different interaction strength parameters @xmath28 and @xmath29 for a nucleon of isospin @xmath30 interacting , respectively , with unlike and like nucleons in the background fields .
more specifically , we use @xmath31 mev and @xmath32 mev . with these parameters ,
the isoscalar potential estimated from @xmath33 agrees reasonably well with predictions from the variational many - body theory @xcite . at the normal nuclear matter density @xmath34 ,
it is consistent with the isoscalar nucleon optical obtained from the nucleon - nucleus scattering experiments @xcite .
the strengthes of the corresponding isovector ( symmetry ) potential can be estimated from @xmath35 . at @xmath34 , by design , the symmetry potential is independent of x and is consistent with the lane potential extracted from the nucleon - nucleus scattering experiments and the ( p , n ) charge exchange reactions @xcite . the corresponding mdi symmetry energy can be written as @xcite @xmath36 - \frac{b x}{\sigma + 1 } \left(\frac{\rho}{\rho_0}\right)^\sigma \nonumber\\ & + & \frac{c_l}{9 \rho_0 \rho } \left(\frac{4 \pi}{h^3}\right)^2 \lambda^2 \left[4 p^4_f - \lambda^2 p^2_f \ln \frac{4 p^2_f + \lambda^2}{\lambda^2}\right ] + \frac{c_u}{9 \rho_0 \rho } \left(\frac{4 \pi}{h^3}\right)^2 \lambda^2 \left[4 p^4_f - p^2_f ( 4 p^2_f + \lambda^2 ) \ln \frac{4 p^2_f + \lambda^2}{\lambda^2}\right],\end{aligned}\ ] ] where @xmath37 is the fermi momentum for symmetric nuclear matter at density @xmath38 . the ibuu04 model can use either the experimental nucleon - nucleon ( nn ) cross sections or the in - medium nn cross sections calculated using an effective - mass scaling model consistent with the single particle potential used @xcite . in the present work the in - medium nn cross sections are used .
the isospin - dependent pauli blocking has been implemented by evolving and checking explicitly neutron and proton phase - space distributions separately .
the coordinates of nucleons in the colliding nuclei are initialized randomly according to the neutron / proton density profiles predicted by the skyrme - hartree - fock approach .
the corresponding fermi momenta are calculated using the local thomas - fermi approximation .
the initial state generated in such way is rather stable for several hundred fm / c without appreciable particle emission in evolving a single nucleus with a momentum - independent mean - field .
however , as it is widely known , see , e.g. , ref , @xcite , momentum dependent mean - field makes the initial state less stable .
for instance , in evolving a single @xmath39sn nucleus with the ibuu04 , at 40 fm / c ( by which most of the nucleons should have freezed - out in heavy - ion reactions at a beam energy of 400 mev / nucleon ) the average ratios of emitted / initial protons and neutrons are 0.6/50 and 1.7/74 , respectively , in calculations using 600 test - particles per nucleon .
the ratios go up to about 1.5/50 and 3.4/74 , respectively , in calculations using 200 test - particles per nucleon .
our results presented in the following are obtained using totally 10,000 events in each case by using 200 test - particles per nucleon in each run of the simulation . because most buu - type transport models including the ibuu04 are incapable of forming dynamically realistic nuclear fragments , some types of after - burners , such as statistical and coalescence models , are normally used as a remedy .
this kind of hybrid models can be used to study reasonably well , for instance , nuclear multifragmentation , see , e.g. , ref.@xcite , collective flow of light fragments , see e.g. , @xcite and the formation of hypernuclei@xcite .
there are , however , some remaining issues , such as the freeze - out time of fragments that is related to the time of coupling the transport model with the after - burner , etc .
there are also interesting work in using advanced coalescence models@xcite , see , e.g. , refs.@xcite .
we notice here that , several advanced cluster recognition routines , such as , the early cluster recognition algorithm ( ecra ) @xcite , the simulated annealing clusterization algorithm ( saca ) @xcite , have been put forward in recent years . for the purposes of the present exploration , however , we use the simplest phase - space coalescence model , see , e.g. , refs.@xcite , where a physical fragment is formed as a cluster of nucleons with relative momenta smaller than @xmath40 and relative distances smaller than @xmath41 .
the results presented in the following are obtained with @xmath42 mev / c and @xmath43 fm .
this simple choice may thus limit the scope and importance of our study here .
for instance , we shall limit ourselves to studies of the relative / differential observables for neutron - proton and t-@xmath0he pairs without attempting to study pairs of the heavier mirror nuclei .
an extended study including the heavy mirror nuclei using the advanced coalescence and/or earlier cluster recognition methods is planned .
noticing that the fopi @xmath3 data favors the symmetry energy with @xmath44 at supra - saturation densities@xcite while the nscl / msu isospin diffusion data favors a symmetry energy at sub - saturation densities between those with @xmath45 and @xmath46@xcite , for comparisons we use here @xmath44 and @xmath46 as two limits of the symmetry energy at high densities .
the corresponding symmetry energy functionals are depicted in the inset of fig .
[ multi ] .
they represent a typically stiff ( @xmath46 ) and a very soft ( @xmath44 ) symmetry energy at supra - saturation densities . in the fopi experiments@xcite ,
the @xmath3 ratio was measured down to a beam energy of 400 mev / nucleon where it shows the largest sensitivity to the high - density behavior of the nuclear symmetry energy@xcite .
the maximum density reached in central au+au reactions at 400 mev / nucleon is about @xmath47 .
it is well known that pions are most abundantly produced in the central collisions . however , the @xmath3 ratio is almost a constant from most central to mid - central impact parameters@xcite .
it is also well known that transverse collective flow is zero in head - on and grazing collisions but is the largest in mid - central collisions .
considering all of the above , as an example , we study @xmath1 reaction at an incident beam energy of 400 mev / nucleon and an impact parameter of 5 fm .
this reaction will be available at fair / gsi in the near future .
we note that the maximum density reached in this reaction is about @xmath48 .
he as a function of the reduced c.m .
rapidity in the reaction of @xmath1 at a beam energy of 400 mev / nucleon and an impact parameter of 5 fm with the symmetry energy parameter @xmath49 ( soft ) and @xmath50 ( stiff ) , respectively.,scaledwidth=50.0% ] before we study the triton-@xmath0he relative and differential flows , it is necessary to first examine the yields of triton and @xmath0he and their ratio .
although these observables and their dependence on the symmetry energy have been studied before especially at lower incident energies , see , e.g. , ref.@xcite , they serve as useful references for measuring the symmetry energy effects on the @xmath3 ratio and the flow observables . shown in fig .
[ multi ] are the rapidity distributions of triton and @xmath4he for the @xmath1 reaction at an incident energy of 400 mev / nucleon and an impact parameter of 5 fm .
it is seen that these clusters are produced mostly at mid - rapidities from the participant region .
the yields , however , are not sensitive to the symmetry energy .
this is not surprising . even at much lower energies where effects of the symmetry energy is stronger , the neutron and proton yields themselves are not so sensitive to the symmetry energy as the yields are dominated by the isoscalar part of the nuclear mean field and nucleon - nucleon collisions . .
the horizonal line at 1.56 ( 2.43 ) in the left ( right ) window is the reaction system s neutron / proton ratio ( squared).,scaledwidth=50.0% ] at fermi energies , the ratio of neutrons to protons or that of mirror nuclei were shown to carry more information about the symmetry energy as effects of the isoscalar potential can be largely cancelled in the ratios@xcite .
however , at significantly higher energies , e.g. , 400 mev / nucleon , except for high transverse momentum particles especially those being squeezed - out perpendicular to the reaction plane@xcite , the neutron to proton ratio becomes less sensitive to the symmetry energy than at lower beam energies@xcite . for comparisons , the t@xmath51he ratio together with the free and bound neutron to proton rations are shown in the left window of fig.[rmulti ] for the @xmath1 reaction . here
, the bound neutrons and protons are the nucleons in the fragments with @xmath52 .
the horizonal line at 1.56 is the reaction system s neutron / proton ratio @xmath53 .
several interesting observations can be made here .
firstly , the @xmath54 ( @xmath55 ) is significantly less ( higher ) than the neutron / proton ratio of the reaction system .
this is the well known isospin fractionation phenomenon@xcite which is reduced here by the production of charged pions .
moreover , at mid - rapidity the @xmath54 shows appreciable sensitivity to the variation of the symmetry energy .
unfortunately , both the @xmath55 and t/@xmath4he ratios show very little sensitivity to the variation of the symmetry energy within statistical error bars , except around the projectile and target rapidities of @xmath56 .
it is especially worth noting that the t@xmath51he ratio is much higher than the free and bound neutron / proton ratios .
a few more comments about this observation are in order here .
first of all , we notice that the assumption of @xmath57he=@xmath55 and that they are equal to the @xmath53 have been widely used in momentum - space coalescence models in the literature , especially at high energies .
our results here and also those reported earlier from transport model simulations@xcite indicate that the coordinate - space correlation is important in considering the cluster formation .
this feature has also been observed in the parton hadronization process in ultra - relativistic heavy - ion collisions by the quark coalescence model@xcite .
in fact , a similarly high @xmath57he ratio has also been observed in the experimental data of heavy - ion reactions at fermi energies , see , e.g. , ref.@xcite .
the possible explanations for the high n / p and t@xmath51he ratios include the isospin fractionation and the overwhelmingly preferential production of symmetric light clusters such as deuteron and alpha particles@xcite in these reactions .
our analyses by turning on / off the coalescence after - burner , the coulomb and symmetry potentials indicate that both mechanism are at work .
it is also interesting to mentioning that , at fermi energies , using a freeze - out temperature extracted from the experiments the rather high t@xmath51he ratio can also be well reproduced within a statistical model@xcite assuming that the difference in the chemical potentials of triton and @xmath0he is dominated by their coulomb potentials@xcite .
the rapidity distributions of the @xmath3 ratio and the bound protons are shown in the right window of fig .
[ rmulti ] .
interestingly , comparing the ratios of all particle - pairs shown in both windows of fig .
[ rmulti ] , it is obvious that the @xmath3 ratio shows the highest sensitivity to the symmetry energy .
more quantitatively , effects of the symmetry energy on the ratio of the total yields of charged pions is about 10% by varying the @xmath14 parameter from @xmath58 to @xmath59 .
this is less than the approximately 20% effect observed in the head - on collisions between two au nuclei at the same beam energy .
this is probably because of the significantly smaller size in @xmath1 although it is more neutron - rich@xcite .
moreover , by comparing the rapidity distributions of the @xmath3 ratio and the bound protons one sees clearly the well - known coulomb focusing effects on the @xmath3 ratio , namely , more @xmath60 s ( @xmath61 s ) are attracted ( repelled ) towards ( away ) from the target and projectile residues@xcite where most of the protons are located in the semi - central reactions considered .
he transverse flows ( in unit of mev ) as functions of the reduced c.m .
rapidity in the same reaction as in fig .
[ multi].,scaledwidth=50.0% ] we now investigate whether the transverse collective flows of triton and @xmath4he can be used to probe the symmetry energy .
firstly , we examine in fig .
[ resflow ] their transverse flows individually .
the average c.m
. transverse momentum per nucleon @xmath62 in the reaction plane is defined as @xmath63 where @xmath64 is the total number of fragments of mass a in the rapidity bin at @xmath65 .
the correlation between the @xmath62 and rapidity @xmath65 reveals the transverse collective flow@xcite .
it is seen that @xmath4he clusters show a stronger flow than triton clusters .
this is mainly due to the stronger coulomb force experienced by the @xmath4he clusters .
more interestingly , the transverse flow of @xmath4he clusters show appreciable sensitivity to the variation of the symmetry energy
. he relative and differential flows ( in unit of mev ) as a function of the reduced c.m .
rapidity in the same reaction as in fig .
[ multi].,scaledwidth=50.0% ] the transverse flow is a result of actions of several factors including the isoscalar , symmetry and coulomb potentials and nucleon - nucleon scatterings .
it is well known that the transverse flow is sensitive to the isoscalar potential .
given the remaining uncertainties associated with the isoscalar potential and the small size of the symmetry energy effects , it would be very difficult to extract any reliable information about the symmetry energy from the individual flows of triton and @xmath4he clusters .
thus techniques of reducing effects of the isoscalar potential while enhancing effects of the isovector potential are very helpful @xcite .
we thus study in fig .
[ clusterflow ] the triton-@xmath4he relative and differential flows .
the relative flow is given as @xmath66 the triton-@xmath4he differential flow reads @xmath67 where @xmath68 , @xmath69 are the number of triton and @xmath4he in the rapidity bin at @xmath65 . from the upper panel of fig.[clusterflow ] , it is seen that the triton-@xmath4he relative flow is very sensitive to the symmetry energy .
because of the larger slope of the @xmath4he flow , the triton-@xmath4he relative flow shows a negative slope at mid - rapidity .
effects of the symmetry energy on the differential flow shown in the lower panel , however , is relatively small .
although the @xmath4he flow is more sensitive to the symmetry energy , the small number of @xmath4he clusters ( as shown in fig . [ multi ] ) makes the @xmath4he flow contributes less to the triton-@xmath4he differential flow ( as indicated in eq .
( [ diflow ] ) ) .
the triton-@xmath4he differential flow is therefore dominated by triton clusters .
consequently , it is less sensitive to the symmetry energy than the triton-@xmath4he relative flow . the slope @xmath70 of the transverse flow at mid - rapidity can be used to characterize more quantitative the symmetry energy effects .
we found that for the t-@xmath0he relative flow , @xmath71 mev / c and @xmath72 mev / c , respectively .
for the t-@xmath0he differential flow , @xmath73 mev / c and @xmath74 mev / c , respectively .
compared to the @xmath3 ratio in the same reaction , the symmetry energy effects on the t-@xmath0he relative and differential flows are much stronger .
thus , especially the t-@xmath0he relative flow can be used as a very useful and independent tool to test the soft symmetry energy at supra - saturation densities extracted from studying the @xmath3 ratio@xcite .
.,scaledwidth=50.0% ] for comparisons , we now study the relative and differential flows for free neutron - proton pairs in fig .
[ nucleonflow ] .
it is seen that they have the same features as the relative and differential flows for triton-@xmath0he pairs .
the larger symmetry energy effects at positive rapidities are due to the more neutron - rich of projectile .
more quantitatively , in terms of the slope parameter @xmath75 , for the n - p relative flow , @xmath76 mev / c and @xmath77 mev / c , respectively . for the n - p differential flow , @xmath78 mev / c and @xmath79 mev / c , respectively . comparing the results in figs.[nucleonflow ] and [ clusterflow ] and
the slope parameters @xmath75 for t-@xmath0he and neutron - proton pairs , we can conclude that they are almost equally useful for probing the density dependence of the nuclear symmetry energy .
in summary , using a hybrid approach coupling the transport model ibuu04 to a phase - space coalescence after - burner we studied the t-@xmath0he relative and differential flows in semi - central @xmath1 reactions at an incident energy of @xmath2 mev / nucleon .
we found that the nuclear symmetry energy affects more strongly the t-@xmath0he relative and differential flows than the @xmath3 ratio in the same reaction .
the t-@xmath0he relative flow can be used as a particular powerful probe of the high - density behavior of the nuclear symmetry energy .
it can be used to test the indications about the symmetry energy at supra - saturation densities observed in the analysis of the @xmath3 data from heavy - ion reactions .
this work was supported in part by the us national science foundation awards phy-0652548 and phy-0757839 , the research corporation under award no.7123 and the texas coordinating board of higher education award no.003565 - 0004 - 2007 , the national natural science foundation of china under grants 10710172 , 10575119 , 10675082 and 10975097 and moe of china under project ncet-05 - 0392 , shanghai rising - star program under grant no.06qa14024 , the srf for rocs , sem of china , and the national basic research program of china ( 973 program ) under contract no.2007cb815004 . | using a transport model coupled with a phase - space coalescence after - burner we study the triton-@xmath0he ratio , relative and differential transverse flows in semi - central @xmath1 reactions at a beam energy of @xmath2 mev / nucleon . the neutron - proton ratio , relative and differential flows are also discussed as a reference .
we find that similar to the neutron - proton pairs the triton-@xmath0he pairs also carry interesting information about the density dependence of the nuclear symmetry energy .
moreover , the nuclear symmetry energy affects more strongly the t-@xmath0he relative and differential flows than the @xmath3 ratio in the same reaction .
the t-@xmath0he relative flow can be used as a particularly powerful probe of the high - density behavior of the nuclear symmetry energy . |
neutron star ( ns ) mergers are strong emitters of gravitational waves ( gws ) and thus among the prime targets for the upcoming gw detectors advanced ligo @xcite and advanced virgo @xcite .
future gw observations of such events @xcite could reveal the properties of high - density matter and nss ( see e.g. @xcite for reviews )
. the merger will likely result in the formation of a differentially rotating , strongly oscillating remnant .
specifically , it has been shown that detections of its _ dominant _ post - merger gw frequency @xmath0 would strongly constrain the radius and the maximum mass of nonrotating nss @xcite .
additionally , there exist potentially detectable _ secondary peaks _ at lower frequencies @xcite , which could improve the constraints on ns properties if the mechanism by which they appear is understood .
in this study we find that _ two different mechanisms _ are at work for producing the low - frequency secondary peaks in the gw spectrum at frequencies below the main peak of the post - merger phase .
specifically , we show that apart from a _ nonlinear combination frequency _ @xcite , there exists also a distinct mechanism that generates a secondary gw peak by the _ rotating pattern of a deformation of spiral shape_. this deformation is initially produced at the time of merging and is then sustained for a few rotational periods .
the consideration of the two different mechanisms leads to a unified classification scheme for the post - merger dynamics and gw emission . for high - mass binaries ( relative to the threshold mass to prompt black - hole collapse ) ,
the nonlinear combination frequency dominates , while for low - mass binaries it is the spiral deformation that produces the strongest secondary peak .
both are simultaneously present and can produce peaks of comparable strength for intermediate binary masses .
hence , the secondary peaks can not simply be explained by the dynamics of the double - core structure alone , which forms the inner part of the remnant , as in @xcite . identifying the importance of the rotating deformation
, we explain for the first time the existence of a low - frequency modulation appearing in addition to the quasi - radial oscillation mode in several physical quantities characterizing the remnant . after clarifying the nature of the secondary peaks in the gw spectrum
, we find mass - dependent empirical relations for the different peaks as a function of ns compactness .
we rule out the existence of a universal , mass - independent and eos independent relation for the secondary peaks , a relation recently proposed in @xcite .
such a relation can only be found within a small sample of eoss for a very limited mass range ( not the same for all eoss ) .
our new relations can be used to optimize the constraints on the eos in the case of the simultaneous detection of several post - merger frequencies .
the understanding of the most prominent features of the gw spectrum is a prerequisite for constructing gw templates for the post - merger phase , which could enhance the detection prospects compared to unmodelled searches @xcite for the advanced ligo and advanced virgo detectors and their discussed upgrades @xcite .
for the planned einstein telescope @xcite direct detections of secondary peaks are a viable prospect @xcite .
we investigate mergers of equal - mass , intrinsically non - spinning nss with a 3d relativistic smoothed particle hydrodynamics ( sph ) code , which imposes the conformal flatness condition on the spatial metric @xcite to solve einstein s field equations and incorporates energy and angular momentum losses by a gw backreaction scheme ( see for details on the code , the setup , resolution tests and model uncertainties ) . comparisons to other numerical setups and also models with an approximate consideration of neutrino effects show an agreement in determining the post - merger spectrum within a few per cent in the peak frequencies @xcite .
magnetic field effects are negligible for not too high initial field strengths @xcite .
we explore a representative sample of ten microphysical , fully temperature - dependent equations of state ( eoss ) ( see table i in @xcite and fig . [
fig : class ] in this work for the mass - radius relations of non - rotating nss of these eoss ) and consider total binary masses @xmath1 between 2.4 @xmath2 and 3.0 @xmath3 . in this work
we consider only nss with an initially irrotational velocity profile because known spin periods in observed ns binaries are slow compared to their orbital motion ( see e.g. @xcite ) , and simulations with initial intrinsic ns spin suggest an impact on the post - merger features of the gw signal only for very fast spins @xcite .
first , we focus on a reference model for the moderately stiff dd2 eos @xcite with an intermediate binary mass of @xmath4 . figure [ fig : spec ] shows the x - polarization of the effective amplitude @xmath5 ( with @xmath6 being the fourier transform of the waveform @xmath7 ) vs. frequency @xmath8 ( reference model in black ) . besides the dominant @xmath9 frequency .
] , there are two secondary peaks at lower frequencies ( @xmath10 and @xmath11 ) with comparable signal - to - noise ratio .
both are generated in the post - merger phase , which can be seen by choosing a time window covering only the post - merger phase for computing the gw spectrum .
the secondary peak shown as @xmath10 is a nonlinear combination frequency between the dominant quadrupolar @xmath9 oscillation and the quasi - radial oscillation of the remnant , as described in @xcite .
we confirm this by performing additional simulations , after adding a quasi - radial density perturbation to the remnant at late times .
the frequency @xmath12 of the strongly excited quasi - radial oscillation is determined by a fourier analysis of the time - evolution of the density or central lapse function and coincides with the frequency difference @xmath13 . as in @xcite , the extracted eigenfunction at @xmath12 confirms the quasi - radial nature .
gw spectra of 1.35 - 1.35 @xmath3 mergers with the dd2 @xcite ( black ) , nl3 @xcite ( blue ) and ls220 @xcite ( red ) eos ( cross polarization along the polar axis at a reference distance of 20 mpc ) . dashed lines show the anticipated unity snr sensitivity curves of advanced ligo @xcite ( red ) and of the einstein telescope @xcite ( black).,width=336 ] the secondary @xmath11 peak is produced by a strong deformation initiated at the time of merging , the pattern of which then rotates ( in the inertial frame ) slower than the inner remnant and lasts for a few rotational periods , while diminishing in amplitude .
figure [ fig : snap ] shows the density evolution in the equatorial plane , in which one can clearly identify the two antipodal bulges of the spiral pattern , which rotate slower than the central parts of the remnant . in this early phase the inner remnant is still composed of two dense cores rotating around each other ( this is the nonlinear generalization of an @xmath14 quadrupole oscillation producing the dominant @xmath15 ) . extracting the rotational motion of the antipodal bulges in our simulations , we indeed find that their frequency equals @xmath16 producing gravitational waves at @xmath17 ( compare the times in the right panels in fig . [
fig : snap ] ; recall the factor two in the frequency of the gw signal compared to the orbital frequency of orbiting point particles ) . in fig .
[ fig : snap ] the antipodal bulges are illustrated by selected fluid elements ( tracers ) , which are shown as black and white dots , while the positions of the individual centers of the double cores are marked by a cross and a circle .
( we define the centers of mass of the double cores by computing the centers of mass of the innermost 1000 sph particles of the respective initial nss and then following their time evolution . )
while in the right panels the antipodal bulges completed approximately one orbit within one millisecond ( @xmath18 ) , the double cores moved further ahead , i.e. with a significantly higher orbital frequency . examining the gw spectrum and considering different time intervals
, we find that the presence of the @xmath11 peak agrees with the appearance and duration of the spiral deformation of the remnant . in the upper right panel of fig .
2 , the spiral deformation can be seen to initially reach deep inside the remnant .
we approximately determine the amount of matter which belongs to the two antipodal bulges that are rotating slower compared to the double cores .
this matter amounts to several tenths of @xmath3 and is thus sufficient to explain the strength of the @xmath11 gw peak .
in addition , we find that the @xmath11 gw peak can be roughly reproduced in a toy model , where the two bulges orbit as point particles around the central double - core structure for a duration of a few milliseconds .
note that this toy model differs significantly from the one in @xcite , which considers only the two cores to be contributing to the gw signal and considers only a single instantaneous orbital frequency of the system .
furthermore , we take advantage of the quadrupole formalism to compute gw spectra considering only certain parts of the remnant , which are defined by using either a density or a spatial cut - off . in fig .
[ fig : rhospec ] we demonstrate that the dominant quadrupole @xmath15 frequency is generated mainly by those regions of the remnant which encompass densities _ exceeding _ 50% of the instantaneous maximum density @xmath19 .
in contrast , most power of the @xmath11 peak originates from densities _ below _
@xmath20 , which corresponds to the outer parts of the remnant , where the two bulges form ( see fig .
[ fig : snap ] ) .
similar conclusions are reached when a spatial cut - off instead of a density cut - off is used .
early post - merger spectrum of 1.35 - 1.35 @xmath3 merger with the dd2 eos ( black ) and gw spectra computed for fluid elements with densities above or below 50% of the maximum density.,width=336 ] in models where @xmath11 dominates over @xmath10 ( see our classification below ) , the presence of the two rotating antipodal bulges explains the appearance of a particular _ low - frequency modulation _ with @xmath21 seen in the time evolution of the central lapse function ( blue curve in fig .
[ fig : lapse ] ) , of @xmath19 , of the size of the central remnant and of the separation between the two cores .
the same modulation occurs as a beat frequency in the time evolution of the gw amplitude ( see e.g. fig . 1 in @xcite ) .
the low - frequency modulation with @xmath21 can coexist with the quasi - radial oscillation , which has a higher frequency @xmath22 .
the above modulation frequency is associated with the rotating bulges and is explained as follows : the central remnant forms an elongated structure , around which the two bulges rotate , lagging behind .
the characteristics of the remnant are modulated depending on the orientation of the antipodal bulges with respect to the double cores : the compactness is smaller , the central lapse function larger and the gw amplitude maximal when the bulges and the cores are aligned ( lower right panel in fig . [
fig : snap ] ) .
instead , when the bulges and the cores are orthogonal to each other ( lower left panel ) , the compactness is largest , the central lapse function smaller and the gw amplitude small .
evolution of the central lapse function for 1.35 - 1.35 @xmath3 mergers with the nl3 ( blue ) and ls220 ( red ) eos ( time - shifted to the same merging time ) .
the evolution of models with an added quasi - radial perturbation is shown ( green ) . for type ii and type iii mergers , like the model with the nl3 eos shown here
, one recognizes a low - frequency oscillation in addition to the quasi - radial mode , which is also strongly excited in the perturbed model at late times.,width=332 ] in a frame corotating with the central remnant ( i.e. a frame rotating with frequency @xmath23 w.r.t .
the inertial frame ) the two bulges counter - rotate with a frequency of @xmath24 which corresponds to the @xmath25 modulation in various quantities ( notice the @xmath26-symmetry of the system , which compensates the factor @xmath27 ) .
our simulations also show that the rotational frequency of the centers of the double - core structure , although not constant , significantly exceeds @xmath16 at any time .
hence , variations in the angular frequency of the double - core structure alone are not sufficient to interpret the secondary peaks , and the simultaneous presence of both @xmath28 and @xmath17 can not simply be attributed to a single instantaneous angular frequency of the system , as suggested in @xcite .
we note that the simultaneous presence of two ( orbital ) frequencies ( of the double - core structure and of the antipodal bulges ) naturally explains strong time variations of the instantaneous gw frequency as seen in many simulations , e.g. @xcite .
also , a single , initially strongly varying instantaneous frequency as explanation for the peaks in the spectrum is incompatible with the fact that the most pronounced peak in the spectrum occurs already at early times .
initially , the instantaneous frequency strongly oscillates around the frequency of the dominant peak .
according to the picture of @xcite , peaks form at frequencies at which the instantaneous frequency spends most time , i.e. at the extrema of the instantaneous frequency .
in this case , one would thus not expect that the dominant @xmath28 is strong at early times .
in contrast , in our simulations the dominant @xmath28 has a substantial strength if one considers only the first few milliseconds .
we have applied the above analysis tools ( gw spectra , determination of @xmath12 from perturbed models , rotational frequency of the antipodal bulges and of the double cores , gw spectra with different cut - off densities ) for a number of representative models , varying the binary mass and stiffness of the eos .
the results fully confirm the generic picture described above . considering this larger set of models , we identify , based on the relative strength between @xmath10 and @xmath11 , three different types of post - merger dynamics and gw spectra for remnants which survive for more than several milliseconds . *
_ type i : _ when the total binary mass @xmath29 is not too far from the threshold mass for prompt quasi - radial collapse of the remnant for a given eos @xcite , the evolution of the central lapse function ( and of @xmath19 ) is dominated by a very strong _ quasi - radial oscillation _ of the remnant , see lower curve in fig .
[ fig : lapse ] . for such models
the two initial nss are more centrally condensed and they merge with higher impact velocity ( fig . 3 in @xcite ) .
because of the strongly excited quasi - radial oscillation , @xmath10 is the strongest secondary peak in the gw spectrum , while @xmath30 is much weaker , likely because for more compact nss the formation of the spiral pattern is less pronounced .
there can be a partial overlap between @xmath10 and @xmath30 , see red curve in fig .
[ fig : spec ] . * _ type ii : _ for intermediate total binary masses , @xmath10 and @xmath11 have a _ comparable strength _ in the gw spectrum and
the two types of secondary peaks are _ well separated _ ( see black curve fig .
1 ) . this is the generic type , which we discussed earlier in our reference model . in characteristic quantities such as the central lapse function
the quasi - radial oscillation frequency @xmath12 as well as the low - frequency modulation with @xmath25 are clearly noticable . *
_ type iii : _ when the total binary mass @xmath29 is significantly below the threshold mass for quasi - radial collapse , the time evolution of the central lapse function ( as well as of @xmath19 and of the radius of the remnant ) is dominated by the @xmath25 modulation that we explained in the previous section as a result of the rotating spiral pattern with the two antipodal bulges . in the evolution of the central lapse function
this modulation typically has a smaller amplitude than type i variations ( fig .
[ fig : lapse ] ) .
the quasi - radial oscillation @xmath12 is also present , but with much smaller amplitude than the dominant modulation because the smaller ns compactness implies a smaller impact velocity and thus a weaker excitation of the quasi - radial oscillation .
the smaller ns compactness also allows for a stronger spiral deformation .
consequently , the dominant secondary peak in the gw spectrum is @xmath11 , while @xmath10 is either very weak or hidden inside the background ( see blue curve in fig .
[ fig : spec ] ) .
different types of post - merger dynamics and gw emission of the different merger models visualized by the mass - radius relations of non - rotating nss of the eoss considered in this work .
the outcome of a given calculation with @xmath1 is shown as symbol at @xmath31 plotted on the mass - radius relation of the eos employed in the simulation .
red squares indicate type i , black crosses stand for type ii , and blue cirlces mark type iii
. see text for definitions of different types of post - merger dynamics and gw emission.,width=336 ] for a given eos there is a continuous transition between the different types of post - merger dynamics depending on the total binary mass .
types i and iii are the limiting cases of the more generic type ii .
notice that since the threshold for quasi - radial collapse is eos - dependent @xcite , the different types cover a different mass range for each eos . for a total binary mass of @xmath32 all three types are possible depending on the eos , where very soft eoss yield type i mergers and very stiff eoss lead to type iii dynamics ( see gw spetra in fig . [ fig : spec ] ) .
this is also shown in fig .
[ fig : class ] , which provides an overview of the types of the different models considered in this study .
the type of a given simulation is indicated by a symbol plotted at the mass of the individual inspiralling stars on the mass - radius relation of the eos which was used in the calculation ( e.g. the results of the 1.2 - 1.2 @xmath3 binaries are displayed at the radii of nss with 1 . 2 @xmath3 ) .
as described , there is a continuous transition between the different classes , which is why one should consider the transitions between the different types in fig .
[ fig : class ] as being tentative and a slightly different classification may be possible at the borders between the different types .
more quantitative definitions of the types may be useful in the future .
still , one can clearly identify a diagonal band of type ii mergers for intermediate binary masses , and also the binary setups leading to the limiting cases of type i or type iii are seen to form roughly diagonal bands . for @xmath33
we find that @xmath11 typically ranges between @xmath34 khz and @xmath35 khz , while @xmath10 ranges between @xmath35 khz and @xmath36 khz .
this property will be useful for identifying either @xmath10 or @xmath11 ( or both ) in future gw observations .
furthermore , we find that @xmath37 _ decreases _ with increasing @xmath29 in all models for which @xmath10 is clearly present , in agreement with the fact that the quasi - radial frequency decreases near the threshold to collapse
. this observation may be useful to estimate the proximity to prompt gravitational collapse . very near the threshold one
thus may expect @xmath38 .
in contrast , @xmath39 typically _ increases _ with increasing @xmath29 , and above the threshold to collapse a spiral pattern during the dynamical collapse could still produce a weak peak in the gw spectrum , as in @xcite .
@xmath28 , @xmath11 and @xmath10 for mergers with ten different eoss and @xmath29=2.7 @xmath3 vs. the compactness @xmath40 for nonrotating , single nss .
solid lines show empirical relations .
the dashed line is taken from @xcite ( see text for explanations).,width=336 ] @xmath11 vs. the compactness , but for different binary masses .
solid lines show empirical relations .
the dashed line is taken from @xcite ( see text for explanations).,width=336 ] @xmath10 vs. the compactness , but for different binary masses .
solid lines show empirical relations .
the dashed line is taken from @xcite ( see text for explanations).,width=336 ]
for our sample of eoss fig .
[ fig : freqa ] shows @xmath28 , @xmath11 and @xmath10 as a function of the compactness @xmath41 of the _ nonspinning , individual nss _
( at infinite separation ) for @xmath4 ( with the compactness in units of @xmath42 ) .
we find strong correlations that can be described by the following quadratic fits : @xmath43&=&199 ( m / r)^2 -28.1(m / r ) + 2.33 , \label{eq : fpeak } \\
f_\mathrm{spiral } [ \rm khz ] & = & 358 ( m / r)^2 - 82.1 ( m / r ) + 6.16 , \label{eq : fspiral } \\ f_{2 - 0 } [ \rm khz ] & = & 392 ( m / r)^2 - 88.3 ( m / r ) + 5.95 .
\label{eq : f20}\end{aligned}\ ] ] the maximum deviations of the data used for these fits are 140 hz , 86 hz and 153 hz for @xmath28 , @xmath11 and @xmath10 , respectively . if the compactness is determined from a measured frequency by inverting eqs . - ,
these maximum deviations imply errors of 3% , 3% and 4% in the compactess for @xmath28 , @xmath11 and @xmath10 , respectively .
( note that the slope in the relation for @xmath11 is somewhat flatter . )
thus , the accuracy of these empirical relations is similarly good for @xmath28 and @xmath11 and only slightly worse for @xmath10 .
even if one assumes an error of 10% uncertainty in the frequency determinations from an actual gw observation , ( which at least for @xmath28 is too pessimistic @xcite ) this would add a 6% , 8% and 5% per cent error in the compactness determination .
similar empirical relations hold , with varying accuracy , for each total binary mass .
as already pointed out and explained for @xmath28 in @xcite , even better empirical relations are obtained for 1.35 - 1.35 @xmath3 binaries when the above three frequencies are examined as function of the radius of a nonrotating ns with 1.6 @xmath2 .
notice that all three empirical relations follow similar trends .
because the frequency of the @xmath11 peak agrees with the rotational frequency of the antipodal bulges near the surface of the remnant , a scaling of @xmath11 with the compactness is not unexpected , which explains the similarities with the behavior of @xmath28 .
in fact , for the models with @xmath4 one finds very tight relations between @xmath11 and @xmath28 , and between @xmath10 and @xmath28 .
a measurement of @xmath28 is already sufficient to accurately constrain the radius of nonrotating nss ( and thus the eos ) @xcite if the total mass has been obtained accurately from the inspiral signal , as it is likely to be the case for distances within which @xmath28 has the required signal - to - noise ratio to be detectable with second - generation interferometers ( e.g. @xcite ) .
a detection of the weaker secondary peaks , with similar @xmath40 dependence as @xmath28 , could further optimize the constraints on the eos .
figure [ fig : freqb ] displays @xmath11 as a function of the compactness @xmath40 of the nonspinning , individual nss for different eoss and for different @xmath29 in the range 2.4 - 3.0 @xmath2 ( the most likely range of total binary masses @xcite ) . the same figure , but for @xmath10
, is shown in fig .
[ fig : freqc ] . especially for @xmath11
we discover that there exist tight relations between the compactness and the secondary frequencies for fixed binary masses . in @xcite the existence of a single , universal , mass - independent relation ( dashed line in figs .
[ fig : freqa ] to [ fig : freqc ] ) between the frequency of the strongest secondary gw peak ( denoted there as @xmath44 ) and @xmath40 was proposed ( there was no distinction of two different secondary peaks , as we find here ) . however , this result was based on using a limited set of five eoss of soft or moderate stiffness ( with corresponding maximum masses of nonrotating ns only up to 2.2 @xmath2 ) as well as on different chosen mass ranges for each eos with a spread of only 0.2 @xmath2 in the total binary mass .
in contrast to @xcite , within our larger sample of ten eoss ( that includes stiff eoss with maximum masses reaching up to 2.8 @xmath2 ) and for a more representative total binary mass range of 2.4 - 3.0 @xmath2 ( same for all eoss ) , _ such a mass - independent , universal relation does not exist _
( dashed curve in figs .
[ fig : freqa ] to [ fig : freqc ] ) . figures [ fig : freqb ] and [ fig : freqc ] show that there is a large spread of up to 500 hz both in the @xmath11 vs. @xmath40 relation and in the @xmath10 vs. @xmath40 relation for an expected range of total binary masses , and that the data can not be described by a single function ( dashed line ) .
even if one consistently chooses the strongest secondary peak among @xmath11 and @xmath10 in each case , there does not exist a mass - independent universal relation with compactness .
the absence of a mass - independent , universal relation implies that the procedure for eos constraints and binary mass determinations as described in @xcite is not valid .
notice also that in fig .
[ fig : freqa ] , the relation proposed in @xcite describes either @xmath11 at low compactness or @xmath10 at high compactness , which is consistent with the expectations from our classification scheme describing which of the secondary peaks dominates the gw spectrum . however , for intermediate values of the compactness the merger will be of type ii and both types of secondary peaks can be present with comparable amplitude , which further complicates the definition of a single @xmath44 frequency , as was assumed in @xcite .
we also point out that detections of the secondary peaks may possibly be less accurate since the peaks are broader in comparison to the main peak and sometimes do not stand out clearly . instead of a universal mass - independent relation , we find that there exist useful empirical relations only for fixed binary masses ( shown as thin line segments in figs .
[ fig : freqb ] and [ fig : freqc ] ) , such as the case shown in fig .
[ fig : freqa ] . for merger events sufficiently close to allow a detection of post - merger gw peaks an accurate determination of the binary masses
is expected ( e.g. @xcite ) . with the upcoming gw detectors
the dominant post - merger oscillation frequency has been shown to be detectable with high precision for mergers within several mpc using a morphology - independent burst analysis @xcite .
significant improvements in the detectability are possible for more sophisticated search algorithms , e.g. matched filtering , which , however , require detailed modelling of the expected signal features to which our present study should contribute by clarifying the relation between dominant and secondary gw peaks .
depending on the highly uncertain binary ns merger rate , a detection may succeed with advanced ligo / virgo @xcite or with discussed upgrades @xcite ( the latter may yield a sensitivity increase at high frequencies of a factor three ) .
since the strength of the secondary features is somewhat lower than that of the dominant peak ( in relation to the anticipated instrument noise curves ) , direct detections of the secondary gw peaks can be expected for the planned einstein telescope @xcite , unless the merger rate is on the more optimistic side as defined in @xcite , which may enable an earlier observation .
the exact detection rate of secondary peaks is hard to quantify not only because of the uncertain merger rate , but also because of the varying strength and prominence of the secondary peaks depending on the exact model . in this study
we focus on equal - mass binaries . to assess the impact of the binary mass ratio , we perform additional simualtions for 1.3 - 1.4 @xmath3 binaries with the dd2 eos @xcite , ls220 eos @xcite and nl3 eos @xcite . the gw spectra in comparison to the symmetric case exhibit qualitatively the same features .
the frequencies of the three different gw peaks are very similar to the equal - mass case with only small deviations of at most 2% , 4% and 6% for @xmath28 , @xmath11 and @xmath10 , respectively , compared to the results of the 1.35 - 1.35 @xmath2 binaries .
also the evolution of the central lapse function shows qualitatively the same behavior .
this suggests that at least for moderately asymmetric binaries our classification scheme applies as well .
it also implies that the frequency dependencies and their implications discussed above hold quantitatively with a good accuracy , although the secondary peaks show somewhat larger deviations , which may impede their use for eos constraints if the mass ratio is not measured accurately from the gw inspiral signal .
we present a unified picture of the post - merger dynamics and gw emission in binary ns mergers . aside from the secondary gw peak produced by a nonlinear coupling between quadrupole and quasi - radial oscillations , we identify a new mechanism
by which another secondary peak in the gw spectrum appears : it is caused by a spiral deformation , the pattern of which rotates slower with respect to the double - core structure of the inner remnant .
based on the presence and strength of these two secondary peaks , we introduce a classification scheme of the post - merger dynamics and of the gw spectrum . moreover , the new mechanism allows for an explanation of low - frequency modulations seen in a number of physical characteristics of the remnant , such as the central lapse function , the maximum density , the separation between the two cores and the gw amplitude .
we find that for fixed total binary masses the frequencies of the secondary gw peaks show a tight relation with the compactness of nonrotating , individual nss .
these relations follow similar trends as the relation between the dominant peak frequency and the ns compactness ( or , equivalently , the ns radius , as shown in @xcite ) . for the dominant peak and for the secondary peak associated with the orbital motion of the antipodal bulges we find the relations between frequency and compactness to be similarly tight .
we rule out the existence of a universal , mass - independent relation for secondary peaks as proposed in @xcite .
identifying the type of a merger event and especially the determination of the quasi - radial frequency by @xmath45 can lead to an estimate of the threshold mass for black - hole collapse and thus to an estimate of the maximum mass of nonrotating nss ( see @xcite ) .
the insights from our classification scheme and from the frequency dependencies found here can provide a basis for constructing future gw template waveforms , increasing the changes for the observation of the post - merger gw emission @xcite .
we will further investigate whether the frequency differences between the dominant and secondary peaks can clarify the nature of a detected secondary peak or if detailed comparisons between spectra of different binary masses are needed .
we will also explore unequal - mass binaries ( anticipating a strong impact of the mass - ratio on @xmath11 ) and analyze the relevance of our classification scheme for the mass ejection and torus formation of ns mergers and for accompanying phenomena , such as r - process nucleosynthesis @xcite , electromagnetic counterparts @xcite and short gamma - ray bursts @xcite .
we thank matthias hempel for providing eos tables , and james clark and thomas janka for helpful discussions .
is a marie curie intra - european fellow within the 7th european community framework programme ( ief 331873 ) .
partial support comes from `` newcompstar '' , cost action mp1304 .
the computations were performed at the rechenzentrum garching of the max - planck - gesellschaft , the max planck institute for astrophysics , and the cyprus institute under the linksceem / cy- tera project . | we introduce a classification scheme of the post - merger dynamics and gravitational - wave emission in binary neutron star mergers , after identifying a new mechanism by which a secondary peak in the gravitational - wave spectrum is produced .
it is caused by a spiral deformation , the pattern of which rotates slower with respect to the double - core structure in the center of the remnant .
this secondary peak is typically well separated in frequency from the secondary peak produced by a nonlinear interaction between a quadrupole and a quasi - radial oscillation .
the new mechanism allows for an explanation of low - frequency modulations seen in a number of physical characteristics of the remnant , such as the central lapse function , the maximum density and the separation between the two cores , but also in the gravitational - wave amplitude .
we find empirical relations for both types of secondary peaks between their gravitational - wave frequency and the compactness of nonrotating individual neutron stars , that exist for fixed total binary masses .
these findings are derived for equal - mass binaries without intrinsic neutron - star spin analyzing hydrodynamical simulations without magnetic field effects .
our classification scheme may form the basis for the construction of detailed gravitational - wave templates of the post - merger phase .
we find that the quasi - radial oscillation frequency of the remnant decreases with the total binary mass . for
a given merger event our classification scheme may allow to determine the proximity of the measured total binary mass to the threshold mass for prompt black - hole formation , which can , in turn , yield an estimate of the maximum neutron - star mass . |
the absolute magnitude of the rr lyrae variables , @xmath9 , is integral to determining distances to old stellar systems in our galaxy and to other nearby galaxies .
for example , rr lyraes are widely used to measure the distances to galactic globular clusters , to the galactic center , and to many members of the local group .
in addition , the distances to individual field rr lyrae stars in the thick disk and halo of our galaxy enable us to determine the kinematic and spatial distributions of these populations .
precise distances to globular clusters are also necessary to determine their ages ( e.g. , buonanno _ et al .
the variation of @xmath9 with abundance , [ fe / h ] , strongly affects the derived age spread and age - metallicity relation of the galactic globular cluster system . these quantities in turn place strong constraints on scenarios describing the formation of the galaxy , specifying the rate of halo formation and whether the chemical enrichment of the halo proceeded in a uniform , global fashion , ( eggen _ et al .
_ 1962 ) or within autonomous star - forming fragments ( searle & zinn 1978 ) .
the zero - point of the @xmath9[fe / h ] relation sets the mean absolute age of the globular cluster system , thus placing a critical lower limit on the age of the universe .
the rr lyraes in a given globular cluster are observed to have a very narrow range of intensity - mean magnitudes , typically @xmath14 = 0.060.15 mag ( sandage 1990a ) .
though the @xmath9 variation from cluster to cluster over a broad range in [ fe / h ] is not so well understood , it is clear that rr lyrae stars have the potential to be excellent standard candles .
historically , there has been discussion over the slope of the @xmath9[fe / h ] relation , with values ranging between @xmath15[fe / h ] = 0 to 0.4 mag dex@xmath5 ( corresponding to a globular cluster age range of @xmath165 gyr to approximately zero ) .
recently , a consensus has begun to form that the slope has a value of 0.150.20 ( e.g. , carney _ et al .
_ 1992 and chaboyer 1995 , though see sandage 1993 and mazzitelli _ et al .
_ 1995 for dissenting opinions ) .
however , the zero - point of the relationship continues to defy a consensus . at the characteristic abundance of the halo , [ fe / h ] = 1.6
, @xmath9 values range from 0.45 to 0.75 mag , and usually fall either toward the brighter or the fainter end of this range .
chaboyer ( 1995 ) showed that this `` two value '' effect translates to a @xmath1722% difference in the derived ages of the galactic globular clusters , and in fact represents the dominant uncertainty in the determination of cluster ages .
a number of methods have been used to estimate @xmath9 , including baade - wesselink ( surface brightness ) analyses ( jones _ et al .
_ 1992 ) , main sequence fitting of globular clusters ( buonanno _ et al .
_ 1989 ) , application of stellar pulsation theory to field stars ( sandage 1990b ) , horizontal branch evolution theory ( lee _ et al .
_ 1990 ) , calibrating the lmc rr lyraes using other lmc distance estimates ( walker 1992 , gould 1995 ) , and the statistical parallax method ( hawley _ et al .
_ 1986 , strugnell _ et al .
_ 1986 ) .
the essence of the statistical parallax ( `` stat-@xmath18 '' ) method is to balance the radial - velocity - derived kinematics of a homogeneous stellar sample with its kinematics as derived from proper motions .
the former are independent of distance , while the latter are distance - dependent .
they are balanced through a simultaneous solution for a distance scale factor .
( 1986 ) discussed the stat-@xmath18 solutions performed prior to 1985 , and described the shortcomings in the methods that were employed .
they argued that only a complete treatment using a maximum - likelihood formulation , together with a minimization technique which is tolerant of inter - dependent variables , can produce accurate solutions .
two modern studies have employed these techniques ( hawley _ et al .
_ 1986 , and strugnell _ et al .
_ 1986 , hereafter referred to as hjbw and srm , respectively ) .
hjbw compiled proper motion , radial velocity , apparent magnitude , and abundance data from the literature to produce a sample of @xmath17140 stars .
srm employed the virtually same set of data .
the two groups obtained similar values for @xmath9 , the only difference being in the details of the adopted reddenings and determination of the apparent magnitudes . both groups concluded that the sample was too small to constrain the slope of the @xmath9[fe / h ] relation .
since then , considerable new data have become available , indicating that it is an opportune time for a new stat-@xmath18 analysis . the lick northern proper motion ( npm )
program ( klemola , jones & hanson 1987 ) has measured absolute proper motions for over 1000 rr lyrae stars .
these data are more uniform and have a more accurate zero - point than the proper motions employed in previous analyses , mainly because the plates , obtained from a single telescope and covering most of the northern hemisphere , were measured and reduced onto a single inertial frame tied to external galaxies . meanwhile , layden ( 1994 ) determined abundances and radial velocities for over 300 rr lyraes , including most of those in the hjbw and srm studies .
blanco ( 1992 ) showed that the abundances used in those studies were of variable accuracy and zero - point .
layden s [ fe / h ] measures are on a self - consistent system and are typically accurate to 0.150.20 dex .
thus , a new stat-@xmath18 solution will reap the benefits of a 50% larger sample size ( 213 stars ) and higher quality data .
furthermore , layden ( 1995 ) showed that the local rr lyrae sample breaks fairly cleanly into thick disk and halo populations at [ fe / h ] = 1 .
the existence of two kinematically distinct populations had not been recognized in previous stat-@xmath18 studies , in part because the existing @xmath19 abundances were of insufficient accuracy to provide the the required abundance resolution . as a result , the two populations had been mixed .
one worries that such mixing might have resulted in the @xmath9[fe / h ] slope being under - estimated . in the worst case ,
population mixing might lead to errant solutions , since the stat-@xmath18 method is a simultaneous solution for kinematics and @xmath9 .
we consider it safest to treat the disk and halo separately in our solutions .
this paper reports the findings of our new stat-@xmath18 solutions , based on the improvements described above . throughout the paper ,
we employ the stat-@xmath18 code used by hjbw . in sec .
2 , we describe in detail the data used in our stat-@xmath18 solutions . in sec . 3 , we begin our analysis with an inverted approach
; we assume an @xmath9[fe / h ] relation , and compute the distance and three space velocity components of each rr lyrae star in our sample .
this gives us insight into how best to divide the sample during the stat-@xmath18 solutions . in sec .
4 , we present monte carlo simulations which enable us to investigate the accuracy of our solutions , and to search for any inherent biases produced by the stat-@xmath18 technique . in sec . 5 , we present the absolute magnitude and kinematic results of the stat-@xmath18 solutions for the observed stars . in sec .
6 , we compare our @xmath9 results with those obtained by other authors using other techniques . in sec
. 7 , we discuss the implications of our results for some basic properties of the galaxy and the universe .
we close with a short summary of our findings .
our primary source of proper motions for this study is the lick northern proper motion ( npm ) program ( klemola , jones & hanson 1987 ) .
the npm program is a photographic survey measuring precise absolute proper motions , on an inertial system defined by 50,000 faint galaxies ( @xmath20 ) , for over 300,000 stars with @xmath21 , covering the northern two - thirds of the sky ( @xmath22 ) , based on plates taken between 1947 and 1988 with the lick 51 cm carnegie double astrograph .
details of the npm observing , plate measurement , and reduction procedures are given by klemola _
et al . _
( 1987 ) .
part i of the npm program , covering the 72% of the northern sky lying outside the heavily obscured regions of the milky way , was completed in 1993 , with the release of the lick npm1 catalog ( klemola , hanson & jones 1993 ; hanson 1993 ) , containing 149,000 stars .
the stellar content of the npm1 catalog is detailed in the npm1 cross - identifications ( klemola , hanson & jones 1994a ; hanson & klemola 1994 ) .
comprehensive error analyses ( klemola , hanson & jones 1994b ) have determined the rms error of the npm absolute proper motions to be @xmath23 in each coordinate , corresponding to a transverse velocity error @xmath24 .
the npm1 catalog contains over 1000 rr lyrae variables ( klemola _ et al .
_ 1994a , appendix 2 ) .
some 300 of these have @xmath25 ( corresponding to @xmath26 kpc ) and may be individually useful for stat-@xmath18 studies .
the particular value of the npm proper motions for this work is that they are on an absolute reference frame .
by contrast , previous rr lyrae stat-@xmath18 solutions have relied on relative proper motions , measured with respect to field stars ( generally with @xmath27 ) , and corrected to absolute by assuming the net motions of the reference stars from models of galactic kinematics and rotation .
using relative proper motions inevitably raises the possibility that the resulting kinematics ( and luminosities ) may depend to some extent on the input assumptions .
specifically , the hjbw and srm stat-@xmath18 solutions used the list of proper motions for 168 rr lyraes compiled by wan , mao & ji ( 1980 ; hereafter referred to as wmj ) .
wmj added new relative proper motions from the shanghai observatory ( wan , he , zhu & li 1979 ) and other sources to the previous such compilation by hemenway ( 1975 ) . because of their absolute character
, we adopted the npm proper motions as the primary source for our stat-@xmath18 database .
our search of the npm1 catalog , using the npm1 cross - identifications , found 171 rr lyrae stars from layden s ( 1994 ) list .
however , the npm1 catalog does not cover low galactic latitudes ( @xmath28 ) , nor the southern sky below @xmath29 declination . to attain the complete sky coverage needed for reliable stat-@xmath18 solutions
, we used the wmj compilation as a secondary source where npm proper motions were not available , adding another 42 stars . using the wmj data raises two practical problems : first , should the wmj motions be corrected to the npm absolute system ?
the heterogeneous nature of the wmj compilation might make any single correction doubtful .
however , clube & dawe ( 1980b ) suggested that the existing rr lyrae proper motions needed a large correction `` in the direction of galactic rotation '' @xmath30 due to incorrect motions of the reference stars , making @xmath31 to 0.1 mag fainter .
so large a systematic error should be easily detectable by comparing the npm and wmj proper motions .
second , how should the wmj data be weighted relative to the npm motions ? wmj estimated the errors for each star from the repeatability of the existing proper motions for that star , but the number of determinations is generally small , so these estimates may not be individually reliable .
for example , the range of @xmath32 is very large , some values are incredibly small ( one is zero ! ) , and the correlation between each star s right ascension and declination proper motion errors is very poor .
one - fourth of the wmj stars have @xmath33 in one coordinate more than three times the value in the other coordinate . because all the sources cited in wmj used methods which should produce errors of equal size in either coordinate ,
this is plainly an artifact of the wmj error analysis .
however , the overall mean and rms errors ( @xmath34 and @xmath35 in @xmath36 ; @xmath37 and @xmath38 in @xmath39 ) are quite comparable to values given in the literature ( wan _ et al . _ 1979 ; hemenway 1975 ) , and may in fact be reliable error estimates . to answer these questions , we compared data for the 109 stars in common between the npm1 catalog and the wmj compilation .
the comparison was done twice ; using @xmath40 in equatorial coordinates ( @xmath41 ) and in galactic coordinates ( @xmath42 ) .
normal probability plots ( lutz & hanson 1992 ) were used for robust estimates of the mean differences and rms dispersions .
plots of @xmath43 vs. @xmath44 and magnitude were examined for any dependence on these observational variables . finally , we assessed the significance of the wmj individual proper motion errors @xmath33(wmj ) .
two principal results were found , bearing on each of the questions posed above .
first , the wmj proper motions in galactic longitude have a small but significant mean difference with npm1 .
we found @xmath45 @xmath46 as clube & dawe ( 1980b ) suggest , such a systematic difference may reflect erroneous reference star motions , but our result is six times smaller than theirs .
consequently , any effects on @xmath9 would be at the 0.01 mag level .
so , there is no need to correct the wmj data for our stat-@xmath18 solutions .
second , the wmj proper motion errors are only useful as an overall average , not on a star - by - star basis . figure 1 shows the npm wmj proper motion differences versus the wmj listed errors . in each coordinate ,
the stars with the smaller errors ( @xmath47 ) scatter just as much as the stars with the larger errors ( @xmath48 ) . for the very smallest errors ( @xmath49 ) the scatter is smaller , but the offsets from zero are large .
a constant error describes the data much better than does the large variation in @xmath32 .
figure 2 quantifies this , testing whether the rms error actually increases with @xmath33 . in each coordinate ,
the 109 stars were sorted by the size of @xmath32 .
figure 2 plots the running values of the rms nominal error and the actual rms dispersion of @xmath43 as a function of this rank .
the nominal error @xmath50 is the quadratic sum of @xmath32 and @xmath51 .
the smooth , slowly rising curve labeled `` tot '' in figure 2 is the running rms value of @xmath50 .
if the wmj errors are correct , then @xmath52 , the rms dispersion of the proper motion differences , should follow this curve .
the jagged line labeled `` @xmath43 '' traces the running value of the actual rms dispersion . in each coordinate
, @xmath52 quickly rises to a large value ( @xmath53 in @xmath36 , @xmath54 in @xmath39 ) by rank @xmath55 , and remains at that level out to rank @xmath56 , rising somewhat at the end .
this proves that the real error of the wmj proper motions is roughly constant , independent of the listed error @xmath32 .
quadratically subtracting @xmath57 from these values gives @xmath58 in ( @xmath59 ) .
almost identical error estimates were obtained from normal probability plots of @xmath60 .
these estimates agree almost perfectly with the rms values of the wmj errors cited above . in view of this result , we gave all wmj proper motions equal weights in our stat-@xmath18 solutions .
test solutions with @xmath61 and @xmath62 , corresponding to the mean and rms wmj errors , respectively , gave very similar results .
we decided to adopt the smaller value , @xmath63 , in order to weight all areas of the sky equally in the stat-@xmath18 solutions ( section 5 ) .
our primary source for rr lyrae metal abundances is the work of layden ( 1994 , hereafter referred to as l94 ) .
these abundances are based on the relative strengths of the k line and the balmer lines h@xmath64 through h@xmath65 , analogous to the @xmath66@xmath67 abundance technique of preston ( 1959 ) .
the abundance scale is tied to the [ fe / h ] abundance scale for globular clusters developed by zinn & west ( 1984 ) , and the individual [ fe / h ] values are typically accurate to 0.150.20 dex .
lambert _ et al_. ( 1995 ) have measured new high - dispersion abundances for a number of the stars in l94 , and find excellent agreement with those results .
jurcsik & kovacs ( 1996 ) also discuss the high quality of the l94 abundances .
previous statistical parallax solutions have used @xmath66@xmath67 values as the metallicity indicator . since then , blanco ( 1992 ) has shown that @xmath66@xmath67 values available to those authors were of variable accuracy and zero - point .
the [ fe / h ] sample of l94 is both self - consistent and contains many more stars than a sample of @xmath66@xmath67 values collected from the literature .
however , there are a few bright rr lyraes in the literature which are not included in the list of l94 . for these
, we adopt the literature @xmath66@xmath67 values , converted to [ fe / h ] using eqn .
we note that some of these values , those taken from hemenway ( 1975 ) , are actually _ inferred _ from photoelectric indices or a period - amplitude-@xmath66@xmath67 relation .
we discuss these stars further in sec .
l94 also measured radial velocities for the stars in his sample , and combined them with literature velocities to produce a catalog of radial velocities , the most accurate currently available , for over 300 nearby rr lyrae stars .
l94 is our primary source of radial velocities .
velocities for a few stars not observed by l94 were taken from the literature compilation shown in table 1 of l94 .
we include these , adopting for their errors the typical errors for each source derived in sec . 2 of l94 .
the existing photometry on field rr lyrae stars is a surprisingly heterogeneous data set .
there are three principal works in the johnson @xmath68-band .
the work of sturch ( 1966 ) is comprised mainly of observations at minimum light ; that of bookmyer _
et al_. ( 1977 ) , which contains as a subset the better - known work of fitch , wisniewski & johnson ( 1966 ) , and is purported to be on the same photometric system ; and that of clube & dawe ( 1980b , hereafter referred to as cd80 ) .
barnes & hawley ( 1986 , hereafter referred to as bh86 ) show that the photometric system of sturch is in good agreement with that of cd80 , and that both are offset from the work of fitch _
et al._. we therefore adopt the cd80 photometry as the standard to which we will compare other photometric works .
many of the existing rr lyrae light curves have incomplete phase coverage , so it is difficult to obtain their intensity - mean apparent magnitudes with accuracy .
however , various relations exist in the literature which allow this quantity to be calculated from light curve extrema , rise times , etc .
( e.g. , fitch _
_ 1966 , cd80 ) .
bh86 recomputed the coefficients for two of these methods using modern , self - consistent data , and find that the method of cd80 gives the tighter relation .
we adopt this parameterization along with their coefficients , @xmath69 where @xmath70 is the intensity - mean magnitude , @xmath71 is the magnitude at minimum light , and @xmath72 is the light curve amplitude . by using the cd80 photometry as the basis of our photometric system , and by employing intensity - mean magnitudes computed from the preceding equation ,
our photometry system is equivalent to that of bh86 .
we computed @xmath70 for the cd80 and bookmyer _
et al_. data sets , and performed a linear regression between them to obtain a transformation between the two photometric data sets ( see table 1 , line 1 ) .
we then adopted data values of cd80 ( 57 stars ) as the primary data , and supplemented it with the values from bookmyer _
et al_. that had been transformed onto the cd80 system ( 81 additional stars ) .
the resulting data set was used to transform the walraven photometry of lub ( 1977 ) onto the cd80 system .
the transformation , given in line 2 of table 1 , is in close agreement with the relation of pel ( 1976 ) .
this resulted in 7 additional stars being added to the data set .
this data set was then used to transform the ccd photometry of schmidt _ et al _ ( 1991 , 1995 ) onto the cd80 system ( see line 3 of table 1 ) , adding 36 stars to the database .
preliminary photometry from layden ( 1996 ; 8 stars ) was also included , though no transformations were possible since there were no stars in common with the database .
data for 24 additional stars were adopted from the photometric compilation of l94 ( his table 9 ) , after converting from the @xmath70 definition of fitch _ et al .
_ 1966 to that of cd80 . clearly , this approach is not ideal , since it relies on the statistical transformations between photometric systems , and assumes that the cd80 system is equivalent to the modern systems used by observers ( e.g. , landolt 1992 ) and theorists ( e.g. , lee _ et al .
the approach has the advantage of reducing systematic errors by placing all the stars on the same photometric system .
ultimately , the transformations result in changes , typically 0.07 mag , which are small compared to the 0.20.3 mag disagreements which arise between different methods of measuring @xmath9 .
furthermore , the sense of the transformation is to brighten the literature values ; had we used the fainter photometry , the @xmath9 values we derived would have been fainter as well . obtaining self - consistent photometry at
the level of several hundredths of a magnitude is a problem shared , but seldom mentioned , by all observers attempting to measure @xmath9 .
srm argued that reddenings derived from the burstein & heiles ( 1982 ) reddening maps provide the most accurate and consistent estimates of rr lyrae reddenings , and hence absorption ( we assume @xmath73 = 3.1 ) .
we therefore use burstein & heiles values when they are available , i.e. , for stars more than 10@xmath74 from the galactic plane .
we reduce their tabulated reddenings by an amount consistent with a uniform dust distribution with an exponential scale height of 100 pc ( see l94 ) . in most cases
this is a small or negligible correction . for stars less than 10@xmath74 from the plane
, we adopt the reddening values given by blanco ( 1992 ) , which are derived from the stars colors at minimum light .
when neither are available , we interpolate between the burstein & heiles reddening at @xmath7510@xmath74 , and that of fitzgerald ( 1968 , 1987 ) at @xmath750.5@xmath74 , at the longitude of the star . the latter is clearly a poor solution , but it is the best available until accurate minimum - light colors can be obtained for these stars .
fortunately , it was used for only 9 stars . using the data sources described above
, we find that 213 stars have values for all five of the fundamental data types : proper motion , abundance , radial velocity , apparent magnitude , and reddening .
this is substantially more than were used in the recent studies of hjbw ( 142 stars ) and srm ( 139 stars ) .
note that we do not consider bailey type-@xmath76 rr lyraes in this study , only type-@xmath77 stars .
this database is presented in table 2 .
the first column gives the variable star name , and the second column gives the npm1 catalog number .
following this are the galactic longitude and latitude ( in degrees ) and the adopted proper motions in right ascension and declination ( in arcsec cen@xmath5 ) .
the seventh and eighth columns give the adopted radial velocity and its error ( in km s@xmath5 ) .
next is the adopted abundance , [ fe / h ] .
the tenth column gives the adopted intensity - mean apparent @xmath68 magnitude , and the eleventh column gives the adopted interstellar absorption .
the twelfth column gives references for the sources of the proper motion , abundance , photometry , and interstellar absorption , as listed at the end of the table .
the final column indicates whether a star was treated as a disk ( 1 ) or halo ( 0 ) star under the three disk / halo definitions discussed in sec .
3 . npm proper motions are used for 171 of the 213 stars in our sample .
abundances from l94 are used for 187 of the stars , and apparent magnitudes computed directly from @xmath68band photometry are used for 182 of the stars . given the dominance of npm proper motions in our catalog , one wonders if the distribution of stars on the sky is skewed , and whether this would introduce a bias into our stat-@xmath18 solutions ( e.g. , croswell , latham & carney 1987 ) . regarding the former , we find that 63% of our sample lies above the celestial equator ; we are weighted to the north celestial hemisphere , but not overwhelmingly so .
similarly , 67% of our stars lie north of the galactic plane
. however , these asymmetries can not produce the kind of biases discussed by croswell _
, since our sample contains no proper motion bias . as klemola _
( 1987 ) describe , samples of `` astrophysically interesting '' stars , such as rr lyraes , were selected for the npm program in advance of the plate measurements , and no star was omitted from the npm1 catalog because its measured proper motion proved to be small .
this type of pre - selection is true of our secondary proper motion source as well . as confirmation
, we note that our data show no sign of the `` croswell effect . ''
the number of stars moving away from the galactic plane is almost identical to the number moving toward it : 106 @xmath78 108 , respectively . finally , we note specifics of interest concerning several stars .
( 1 ) bx dra was shown by schmidt _
( 1995 ) to be an eclipsing binary rather than an rr lyrae ( kholopov 1985 ) ; it was removed from our database .
( 2 ) l94 found sv boo to have [ fe / h ] = 0.43 from a single low - quality spectrum , whereas hemenway ( 1975 ) quoted @xmath66@xmath67 = 7 ( [ fe / h]@xmath79 = 1.55 ) . since the kinematics of sv boo suggest it belongs to the halo , we adopt the hemenway abundance .
( 3 ) the radial velocity of bb pup was revised to 98 @xmath75 9 km s@xmath5 from that in l94 by eliminating the outlier velocity 255 @xmath75 16 km s@xmath5 from the list in table 2 of l94 .
( 4 ) for three stars , ae dra , bd dra , and bk eri , the proper motion was improved by removing one discordant measurement from the average value quoted in the npm1 catalog .
( 5 ) the seven stars listed with the abundance source `` 3 '' in table 2 all had photometrically - determined [ fe / h ] values from hemenway ( 1975 ; see sec .
2.2 ) which suggested that they were thick disk stars .
however , their kinematics suggested that all seven stars are halo members .
we therefore set [ fe / h ] = 1.5 for these stars .
( 6 ) we found that rx cvn historically has been misidentified . the proper motion of wmj and the radial velocity of joy ( 1950 ) give a space velocity greater than the escape velocity of the galaxy .
the npm proper motion and the radial velocity measured by l94 ( see our table 2 ) give a reasonable space velocity for a halo star
. inspection of the npm plates shows that the measured star has a nearby companion , probably a foreground dwarf , which was probably mistakenly observed by wmj and joy .
the rr lyrae is the eastern - most of the pair .
layden ( 1995 , hereafter referred to as l95 ) showed that the rr lyraes separate into a halo and a ( primarily thick ) disk population at [ fe / h ] = 1.0 .
we wish to see if this separation persists using our improved database
. to do this , we computed provisional distances to the stars in our sample using the @xmath9[fe / h ] relationship of carney , storm & jones ( 1992 , hereafter referred to as csj ) , @xmath9 = 0.15[fe / h ] + 1.01 mag .
we then computed the stars @xmath80 , @xmath68 , @xmath81 space velocities , following johnson & soderblom ( 1987 ) with one exception : we take @xmath80 as positive outward . at distances of 12 kpc , typical of the distances in our sample , the sun - oriented @xmath80 , @xmath68
, @xmath81 frame can be misaligned by a small angle in the @xmath80 , @xmath68 plane ( @xmath82 deg ) from the cylindrical galactic directions ( @xmath18 , @xmath83 , @xmath84 ) at each star .
so , we rotated the @xmath80 , @xmath68 , @xmath81 velocities into the @xmath18 , @xmath83 , @xmath84 frame , after adding to @xmath68 the iau standard rotational velocity @xmath85 km s@xmath5 ( kerr & lynden - bell 1986 ) , and after adding the `` dynamical '' solar motion ( 9 , + 12 , + 7 km s@xmath5 ; mihalas & binney 1981 , p.400 ) to correct @xmath80 , @xmath68 , @xmath81 to the local standard of rest ( lsr ) .
the results are the velocity components @xmath86 , @xmath87 , and @xmath88 , where @xmath86 increases outwards from the axis of galactic rotation , @xmath87 increases in the direction of galactic rotation , and @xmath88 increases toward the north galactic pole .
figure 3 shows @xmath86 and @xmath88 plotted against @xmath87 . clearly , the stars with [ fe / h ] @xmath89 1.0 have large velocity dispersions and little net galactic rotation , typical of the halo .
meanwhile , the stars with [ fe / h ] @xmath90 are clustered around @xmath87 @xmath91 km s@xmath5 .
figure 4 shows @xmath87 as a function of abundance .
while the distribution of stars in figures 3 and 4 is consistent with the first - order view of a disk / halo separation at [ fe / h ] = 1.0 , there are four stars with @xmath87 @xmath92 km s@xmath5 at [ fe / h ] @xmath93 , whose extreme kinematics clearly mark them as members of the halo . similarly , there may be an excess of stars with @xmath87 @xmath94 km s@xmath5 at [ fe / h ] @xmath95 , which may belong to the `` metal - weak thick disk '' ( mwtd ; morrison , flynn & freeman 1990 ) .
l95 discussed the presence of such stars in his sample of rr lyraes .
it is not possible to assign these stars individually to the disk or halo populations without ambiguity .
we therefore separate the disk from the halo using three distinct definitions , and perform the stat-@xmath18 solutions for each set of definitions , in order to test the effects of the different definitions on the derived kinematics and absolute magnitudes .
the three disk / halo separation definitions are summarized in table 3 .
all three definitions assign the four low-@xmath87 stars with [ fe / h ] @xmath90 to the halo .
the first definition , similar to that of nissen & schuster ( 1991 ) , admits a small number of mwtd rr lyraes , primarily with @xmath96 [ fe / h ] @xmath95 , in agreement with l95 .
the second assumes that no mwtd ( [ fe / h ] @xmath95 ) rr lyraes exist .
the third definition admits a larger population of mwtd rr lyraes , which reaches to [ fe / h ] @xmath97 , more along the lines of the population of red giants described by morrison _ et al_. ( 1990 ) .
two stars , ao peg and fu vir , fit one or more of the disk definitions , yet clearly belong to the halo based on their extreme kinematics : ( @xmath98 ) = ( 212 , + 236 , 207 ) and ( 178 , + 249 , 93 ) , respectively .
we have moved these stars into the corresponding halo definitions , as noted in table 3 .
we note that the choice of @xmath9[fe / h ] relations used to compute the distances does not significantly affect the separation .
only one star crosses the sloping disk / halo line in fig .
4 when we change from the csj @xmath9[fe / h ] relation to that advocated by sandage ( 1993 ) .
in sec . 5 we will present the stat-@xmath18 analysis of our data . first , however , we will test the hjbw stat-@xmath18 algorithm itself , using synthetic data with known properties ( positions , velocities , @xmath9 , etc . ) .
this step seems vital to ensure reliable results .
specifically , we test : ( 1 ) how accurately the stat-@xmath18 solutions reproduce the kinematics and luminosities of the input data ; ( 2 ) how reliable the error estimates are ; ( 3 ) the sensitivity to the number of stars and their distribution on the sky ; ( 4 ) whether all the free parameters in the hjbw stat-@xmath18 algorithm are necessary ; ( 5 ) whether the small misalignment ( sec .
3 ) between @xmath99 and @xmath100 has any significant effects ; ( 6 ) whether the results are biased by any input assumptions ; and ( 7 ) whether any corrections are necessary for bias in the results .
the hjbw algorithm uses a simplex optimization technique to maximize the likelihood in murray s ( 1983 ) kinematic model ( also used by srm ) .
there are 11 free parameters : the solar motion ( @xmath101 ) , the velocity ellipsoid ( @xmath102 ) with three covariances ( @xmath103 ) to allow an arbitrary orientation , the distance scale parameter @xmath104 and its dispersion @xmath105 .
the solution also returns an approximate standard error estimate @xmath106 for each parameter .
the solution returns an absolute magnitude @xmath107 by differential correction to a starting value @xmath108 , which for convenience we fix at + 1.0 .
we refer the reader to hjbw for full details of the stat-@xmath18 algorithm .
we made two modifications to the hjbw procedure : ( 1 ) we fixed an error which caused the large velocity dispersion errors listed in table 2 of hjbw ; and ( 2 ) when necessary , we inspected the data to reject extreme outliers and repeated the solutions . to perform monte carlo tests we generated an ensemble of simulated data sets as outlined in table 4 .
these were designed to realistically simulate the halo and disk subsamples of our real data ( sec .
3 ) . for the halo and disk simulations h1 and d1 ,
@xmath109 were randomly assigned spatial positions ( @xmath110 ) and [ fe / h ] values from uniform distributions with appropriate limits .
@xmath107 values were then computed from the @xmath9[fe / h ] relation of csj , with a gaussian cosmic dispersion @xmath111 mag .
space velocities ( @xmath86 , @xmath87 , @xmath112 ) were generated from a gaussian velocity ellipsoid whose parameters are given in table 4 .
each star s right ascension , declination , proper motion , radial velocity , and apparent magnitude were then calculated , with gaussian observational errors 0.5 arcsec cen@xmath5 and 20 km s@xmath5 added to the proper motion components and the radial velocity , respectively . for each simulation ( h1 , d1 ) we created @xmath113 different data sets ;
@xmath114 was set larger for the disk simulations because @xmath115 is proportionately smaller . as discussed in sec .
2 , the actual distribution of our rr lyraes on the sky is far from uniform . to test whether this affects the stat-@xmath18 results , we prepared alternate data sets ( h2 , d2 ) using the observed sky positions , apparent magnitudes , and metallicities from table 2 , with the disk-1/halo-1 separation of table 3 .
then , @xmath114 sets of synthetic proper motions and radial velocities were randomly generated as above .
for each data set ( h1 , h2 , d1 , d2 ) , stat-@xmath18 solutions were performed as outlined in column 6 of table 4 .
multiple solution sets tested particular parameters in the hjbw model . to test the effect of the hjbw distance scale dispersion parameter @xmath105 we ran two solutions for each halo data set , with @xmath116 as indicated in column 7 of table 4 .
these solutions will be discussed in sec .
4.6 . for the data sets h2 and
d2 we ran solutions with and without the velocity covariance parameters ( @xmath117 ) , as indicated in column 8 of table 4 .
the reasons for this will be discussed in sec .
fifteen additional h2 data sets ( row 5 of table 4 ) were created in order to test with higher precision the important solution sets which excluded the velocity covariance parameters .
the significance of these solutions will be discussed in sec .
finally , we note that for a few of the disk data sets , the solutions failed to converge ( @xmath118 ) .
this will be discussed in sec .
before discussing the stat-@xmath18 solutions , we need to deal with an apparent inconsistency between the data simulations and the hjbw solution method .
we generated our simulated velocities using velocity ellipsoids oriented to the cylindrical coordinate frame ( @xmath86 , @xmath87 , @xmath112 ) which is physically the most appropriate frame for galactic velocities .
however , the hjbw algorithm uses the ( @xmath101 ) rectangular coordinate frame to compute the solar motion and stellar velocity dispersions . in sec . 3
, we noted that the ( @xmath119 ) coordinates can be misaligned with ( @xmath86 , @xmath87 , @xmath112 ) , by a small angle ( rms @xmath120 deg ) .
it is important to show that any effects of this misalignment are too small to significantly affect our results .
we can do this directly with the simulated data , because for every star we know the space velocity in each coordinate frame .
this lets us measure , for every data set , any differences ( 1 ) between @xmath121 and @xmath122 and ( 2 ) between @xmath123 and @xmath124 . in the remainder of this paper
, we refer to the values computed directly from the simulated data as the `` true '' values for the data set . for the halo ( h1 , h2 ) ,
the principal effect is the projection of the long axis of the velocity ellipsoid ( @xmath125 ) partly onto the @xmath68 axis , increasing @xmath14 at the expense of @xmath126 .
quantitatively , @xmath127 km s@xmath5 , and @xmath128 km s@xmath5 . clearly , these effects are far too small to be of any concern here .
for the disk ( d1 , d2 ) , the principal effect of coordinate misalignment is the partial projection of the rotation vector @xmath129 km s@xmath5 onto the @xmath80 axis .
distant stars in the direction of galactic rotation @xmath130 get a negative contribution to their @xmath80 velocity ; stars toward ( 270,0 ) get a positive contribution .
the net effect is that @xmath131 km s@xmath5 .
also , @xmath132 is decreased by 1 km s@xmath5 .
again , these effects are considerably smaller than the observational errors , and can be neglected in practice .
stat-@xmath18 solutions for each of the 25 halo and 40 disk data sets were performed as outlined in table 4 .
the results for each set of solutions are summarized in table 5 .
the left side of table 5 gives results for the @xmath68 component of the solar motion , the three velocity dispersions ( @xmath133 ) , and @xmath107 .
the @xmath66 values on the right side of table 5 represent the differences ( solution @xmath134 `` true '' ) for each quantity .
table 5 omits the @xmath80 and @xmath81 solar motion components , as these were always equal to the true values to @xmath135 km s@xmath5 .
each solution set in table 5 lists two rows of results .
the first row gives the mean of each quantity .
the second row gives two different estimates of the uncertainty in the quantity : @xmath136 ( left side of table 5 ) gives the mean of the internal error estimates returned by the hjbw code for each parameter , while sd ( right side ) gives the rms dispersion about each mean difference .
the latter external error estimates reflect how precisely the stat-@xmath18 program returns the `` true '' values .
the values in row 5 of table 5 are more precise than those quoted for the other halo simulations , since they are based on 20 rather than 5 data sets .
the major result of the monte carlo simulations is that the hjbw algorithm does an excellent job of returning the `` true '' input parameters , to within a few km s@xmath5 for the velocities and dispersions , and to within @xmath137 mag for @xmath107 .
this is true for both the halo and the disk simulations , and for both the random and real space distributions .
furthermore , in nearly all cases the external errors ( sd s of the @xmath66 s ) are no larger than the hjbw program s internal error estimates @xmath138 , and in some cases they are considerably smaller .
detailed examination of table 5 shows some small systematic effects which are worth considering further .
note that 7 out of 8 @xmath139 values are positive , all 24 @xmath140 values are negative , and 7 out of 8 @xmath141 values are positive .
examination of the individual solutions shows that these same effects occur on a solution - by - solution basis , with @xmath68 , @xmath123 , and @xmath107 all varying in tandem , linked together by the hjbw distance scale parameter @xmath104 .
the average values are slightly biased toward a `` short '' distance scale ( @xmath142 ) .
this bias is larger for the real space distribution data sets ( h2 , d2 ) than for the random sets ( h1 , d1 ) . whether we can or should correct for this small bias will be discussed in sec .
4.8 . for the halo ( h ) solutions , the internal error estimates @xmath143 returned by the hjbw program tend to be larger than the external errors
for example , @xmath138 for @xmath107 is 0.12 mag , while the sd of @xmath144 averages 0.08 mag . for the kinematic parameters , @xmath145/sd @xmath146 .
these results indicate that the real accuracy of the halo solutions may be better than the internal errors claim .
however , given the simplified nature of the simulations , we will conservatively adopt the internal error estimates in discussing our real data solutions ( sec .
5 ) . for the disk ( d ) solutions ,
the internal and external errors for @xmath107 generally agree , though there is a small discrepancy for the kinematic parameters , @xmath145/sd @xmath147 .
in addition , the standard error of @xmath107 is @xmath170.3 mag , 34 times as large as for the halo .
the failure of the @xmath107 errors to follow an @xmath148 relation may mean that @xmath149 is near the lower limit for successful solutions ( see sec .
4.5 ) , but it must also reflect the fact that the stat-@xmath18 method inherently works better for a population with a larger velocity dispersion . the test solutions ( h1.0 , h2.0 ) with the hjbw distance scale dispersion parameter @xmath105 set to zero will be discussed in section 4.6 .
the test solutions ( h2d , d2d ) without the velocity ellipsoid covariances ( @xmath103 ) will be discussed in section 4.7 . since the number of stars in our disk samples ( both real and synthetic data ) is near or below the lower limit ( 50 stars ) that hjbw found necessary for successful solutions , we need to ask how reliable our disk solutions can be with so few stars .
the symptom of having too few stars is that the hjbw likelihood function becomes ill - conditioned , and the iterative solution fails to converge to finite output parameters ( hence @xmath118 in table 4 ) .
surprisingly , given the hjbw result , over 90% of our monte carlo disk solutions , and all 3 real - data disk solutions ( sec .
5 ) did in fact converge .
moreover , table 5 shows that the disk output parameters are well - determined , i.e. , they have little bias and accurate error estimates .
the convergence of our monte carlo solutions for @xmath150 may reflect the gaussian nature of our simulations , in contrast to the vagaries of the real data that hjbw used . in our experience , the hjbw algorithm is not resistant to large outliers ; for real data , this makes the disk / halo separation quite critical .
the success of our real - data disk solutions for @xmath151 is most likely due to the better separation we achieved using [ fe / h ] and @xmath152 ( sec .
4 ) instead of @xmath19 and period ( hjbw ) . without doing many more simulations ,
it is not possible for us to state what the true lower limit on @xmath115 may be . nor is this necessary , since the clear result of our monte carlo simulations is that solutions that do converge give reliable results .
both hjbw and srm found a strong correlation in their stat-@xmath18 solutions between @xmath9 and @xmath153 , the cosmic dispersion in @xmath9 . in the murray ( 1983 )
model , the cosmic dispersion is parameterized by @xmath105 , the dispersion in the distance scale parameter @xmath104 .
this correlation effectively prevents solving for @xmath105 ; instead this parameter must be fixed at a value chosen to represent a reasonable value of @xmath153 .
equation 8 of hjbw relates @xmath153 , @xmath105 , and @xmath104 . for @xmath154 , @xmath155 . for @xmath156 and @xmath157 from our solutions , @xmath158 mag .
observations show this to be a reasonable range of @xmath153 .
sandage ( 1990a ) found the intrinsic dispersion of rr lyrae magnitudes within a globular cluster ( i.e. , at a single metallicity ) to be @xmath14 = 0.060.15 mag .
for our field rr lyraes there will be an additional dispersion @xmath159 proportional to the slope of the @xmath9[fe / h ] relation .
we estimated @xmath159 numerically by populating various @xmath9[fe / h ] relations with `` stars '' having the [ fe / h ] distribution of our halo sample .
the total @xmath153 is then the quadratic sum of @xmath14 and @xmath159 . for @xmath160 mag , using the csj @xmath9[fe / h ] relation ( slope = + 0.15 mag dex@xmath5 ) we obtain @xmath161 mag .
using the steeper slope ( + 0.39 ) advocated by sandage ( 1990b ) gives @xmath162 mag .
to reach @xmath163 mag , we must adopt an extreme value of @xmath164 mag along with the sandage ( 1990b ) slope .
consequently , a reasonable range of @xmath105 to test in our monte carlo solutions is @xmath165 .
solution sets ( h1.0 , h1 ) and ( h2.0 , h2 ) with @xmath166 respectively ( table 4 ) let us evaluate the effects on @xmath9 . the results in table 5 indicate that @xmath9 comes out @xmath170.03 mag brighter for @xmath156 than for @xmath167 .
similar results were found by hjbw and srm .
table 5 shows that the assumed value of @xmath105 does not affect the derived kinematics .
clearly , our choice of @xmath105 will only have a small effect on our stat-@xmath18 results . to be conservative ,
we will adopt @xmath156 to analyze our real data ( sec .
5 ) . for a given solution the true value of @xmath9
thus may be a few hundredths of a magnitude fainter than the value we derive .
three of the 11 parameters in the murray ( 1983 ) model are the covariances ( @xmath168 which allow the velocity ellipsoid to have an arbitrary orientation with respect to the principal galactic directions .
the results of hjbw strongly suggest that this may not be necessary in practice .
given the apparent benefits of eliminating unneeded parameters from the model ( especially for the disk , where @xmath169 does not greatly exceed the number of free parameters ) , it seems wise to use our monte carlo simulations to test whether the covariances ( @xmath103 ) are needed in our real - data stat-@xmath18 solutions .
because our simulated data were generated with no correlations among the ( @xmath101 ) velocities , the covariances ( @xmath117 ) returned by the stat-@xmath18 solutions simply reflect random scatter in the data .
thus the covariances for the real - data solutions can be tested for statistical significance by comparison with the monte carlo simulations . to do this , for each stat-@xmath18 solution we calculated the correlation coefficients ( cc s ) @xmath170 , etc .
for each of the sets of simulated data listed in table 4 , we computed the mean , sd , and range of the cc s of the individual trials .
for all the simulations , the mean correlations were near zero , as expected .
for the halo , ( both h1 and h2 ) each of the three sd s was @xmath171 . by comparison ,
the rms value of the real - data halo cc s was 0.10 , with none of the cc s exceeding the range of the simulated values .
for the disk , the sd s were @xmath172 for set d1 ( random space distribution ) , and @xmath173 for set d2 ( real space distribution ) .
the rms value of the real - data disk cc s was 0.13 ; again none of the cc s exceeded the range of the simulated values . from these tests
we conclude that there is no evidence , for either the halo or the disk rr lyraes , that the velocity ellipsoid deviates from the principal directions ( @xmath101 ) .
consequently , we can run our stat-@xmath18 solutions with the covariance parameters ( @xmath174 ) removed from the model .
the entries in table 5 for solution sets h2d and d2d show the results of these solutions on the data sets h2 and d2 .
there is little difference in the kinematics or absolute magnitudes produced by the new solutions , save for a slight tendency for the errors to be reduced . analyzing our real data ( sec .
5 ) without the unneeded covariances reduced the disk @xmath107 errors by 10% . in sec .
4.4 we found that the results in table 5 suggested that the stat-@xmath18 solutions may be slightly biased toward a `` short '' distance scale .
the solution velocities are consistently several per cent too small , and @xmath175 mag too faint .
the bias was largest ( @xmath176 mag ) for the data set d2 , the disk simulation with the space distribution of the real rr lyraes .
although this bias is much less than the random error ( @xmath177 mag ) of a single disk solution , it may still be worth applying corrections to our real - data results ( sec .
the above - mentioned bias toward a `` short '' distance scale should not be confused with a bias toward the _ fainter _ @xmath107 regime discussed in secs . 1 , 6 , and 7 ( i.e. , toward @xmath178 mag at [ fe / h ] @xmath97 , rather than + 0.4 mag ) .
it can easily be demonstrated that our solutions have no intrinsic bias toward any particular @xmath107 value ; nor are the solutions biased by the choice of the starting value @xmath108 .
tests using simulated data sets based on @xmath9 0.5 mag brighter than the csj relation adopted in sec .
4.2 correctly returned @xmath107 values 0.5 mag brighter , with unchanged kinematics .
these results were recovered exactly when a brighter starting value @xmath179 was used , and to within @xmath180 mag in @xmath107 and @xmath181% in the velocities even when the usual @xmath182 was used . in the spirit of exploratory data analysis , we plotted @xmath144 versus the `` solution '' and `` true '' values of the velocities ( @xmath183 and dispersions @xmath123 for all 65 simulations .
this led to the discovery that , for the real space distribution sets ( h2 , d2 ) , the bias was smallest when the velocity ellipsoid was `` long '' , i.e. when @xmath184 , and largest when the velocity ellipsoid was `` round '' , i.e. when @xmath185 . thus the disk is chiefly affected , while the halo is not .
the 15 additional h2d solutions outlined in table 4 were performed to illuminate this situation .
figure 5 plots @xmath144 as a function of @xmath186 for solution sets h2d and d2d .
because of their markedly different distributions on the sky , we consider the halo and disk separately in fig .
5 . for the halo , the small bias ( @xmath187 mag ) is well - determined because of the increased number of h2d solutions , but the slope is not statistically significant . for the disk , both
the mean bias ( + 0.13 mag ) and the slope ( 0.69 mag ) are 2-@xmath188 significant .
we suggest that these results may be used to apply a bias correction ( subtracted from @xmath107 ) to our real - data results ( sec .
5 ) for the disk , as a function of @xmath189 . ( we choose not to apply a @xmath144 correction to the halo results for two reasons
. the halo correction would be small compared to the other sources of error in @xmath9 discussed in sec .
6 ; moreover the bias is compensated by the roughly equal , but opposite , @xmath105 effect discussed in sec .
4.6 . ) for consistency , when we apply the disk @xmath144 correction in sec . 5
, we will also correct the disk kinematics for the `` short '' distance scale by enlarging the velocities and dispersions by a factor of @xmath190 .
two objections may be raised to these corrections : first , that the reason for the bias is not understood , and second , that the `` true '' value of @xmath189 is not known for the real data
. the latter problem can be overcome by computing @xmath126 and @xmath14 directly from the data , as in sec .
3 . because we only need the ratio @xmath189 , any distance dependence from the assumed @xmath9[fe / h ] relation cancels out .
it remains mysterious to us why the disk stat-@xmath18 solutions are slightly biased when the velocity ellipsoid is `` round '' .
since the effect occurs for the real ( but not a random ) stellar distribution , it must be caused by the the uneven distribution of the disk rr lyraes on the sky .
it might be possible to solve this puzzle with a much larger set of simulations , but that is clearly beyond the scope of this paper .
we note that maximum - likelihood methods in general are not unbiased ; clube & dawe ( 1980a ) found an equal @xmath107 bias ( @xmath191 mag ) in the opposite direction !
we conclude simply that since our disk @xmath107 bias is relatively large and can be calibrated as a function of the observational variables , we should apply it to our real - data solutions .
applying the lessons learned from our monte carlo simulations ( sec .
4 ) , we analyzed our real data ( each of the disk / halo subsamples in table 3 ) by running the stat-@xmath18 program with @xmath156 and performing two sets of solutions , with and without the velocity ellipsoid covariances ( @xmath103 ) .
as in sec .
4.7 , the differences between the two sets of solutions were small .
the kinematics generally changed by @xmath89 1 km s@xmath5 ; @xmath107 averaged @xmath192 mag brighter in the solutions without the covariance terms .
most important , the errors returned by the stat-@xmath18 program for the disk data sets were typically 10% smaller without the covariances , presumably owing to the larger number of degrees of freedom attained by removing three free parameters from the solutions . since the correlation coefficients ( @xmath193 ) did not prove to be significant ( sec .
4.7 ) , we therefore adopt these solutions , with the velocity ellipsoid aligned to the principal galactic directions ( @xmath101 ) , as our final results .
table 6 presents , for each solution as discussed below , the solar motion @xmath194 , the velocity ellipsoid @xmath123 , and the absolute magnitudes . below each of these entries
is presented the standard error for that term , computed by the stat-@xmath18 program .
recall that the monte carlo simulations suggested that the true uncertainties may be as much as 2 times _ smaller _ than those quoted in the table .
the final column of table 6 shows the @xmath9 values after correction for the @xmath195 bias discussed in sec .
examination of table 6 shows that the kinematics of the rr lyraes do not depend significantly on which disk / halo definition is employed . because definition 3 of table 3 gives the purest halo sample and the largest , best - determined disk sample , we adopt halo-3 and disk-3 as our best solutions in table 6 . thus , our best estimates of the net rotation and velocity ellipsoid of the halo rr lyraes are @xmath196 @xmath197 for the disk rr lyraes , after correcting ( by a factor of 1.07 ) for the distance scale bias ( @xmath198 mag , sec .
5.2 ) , we obtain @xmath199 @xmath200 we note ( see sec .
4.4 ) that the hjbw stat-@xmath18 program implicitly accounts for the effects of observational errors in the determination of the velocity dispersion parameters , so these results are unbiased estimates of the true velocity dispersions of the halo and disk rr lyrae populations .
these kinematic values are in excellent agreement with the rr lyrae kinematics derived by l95 .
they also correspond quite well with the kinematics of the thick disk and halo based on other tracer populations ( _ cf .
_ casertano _ et al .
_ 1990 ; l95 table 8) , with two possible exceptions .
first , the vertical velocity dispersion of the disk , @xmath201 km s@xmath5 , is somewhat smaller than the typically quoted value of 3545 km s@xmath5 .
l95 suggested that the rr lyrae `` disk '' subsample contains stars from both the thick disk and the old thin disk populations , such that the net kinematics are intermediate between the two .
unfortunately , the disk sample contains too few stars for us to subdivide it and perform meaningful stat-@xmath18 solutions for separate thin and thick disk components .
second , @xmath126 for the halo rr lyraes is large compared to many other estimates , 168 km s@xmath5 @xmath78 120155 km s@xmath5 .
the cause of this effect is less clear
. it may be due to our removing interloper thick disk stars more completely than other studies ( l95 ) , or it may be related to a subtle selection bias experienced by rr lyraes .
for example , if the halo is composed of an accreted component and a dissipatively - formed component ( zinn 1993 , majewski 1993 ) , if the components have different kinematics ( e.g. , beers 1996 ) , and if rr lyraes are more easily formed in one component than the other , then using rr lyraes as kinematic tracers would bias the kinematic results to favor one or the other halo components , relative to their representation in samples using other stellar tracers . at present , it seems that the halo rr lyraes may be preferentially tracing the accreted halo component , though a detailed analysis outside the scope of this paper is required to further address this problem . in table 6 , @xmath107 is virtually the same for each of the three halo definitions .
as above , we adopt halo-3 as our purest definition of the halo rr lyraes .
this gives @xmath202 } \rangle = -1.61.\ ] ] as discussed in secs .
4.8 and 6 , no bias corrections have been applied to the halo solutions . for the disk ,
@xmath107 is more sensitive to which set of stars is used in the stat-@xmath18 solution . as in the monte carlo simulations ( sec 4.4 )
, the errors in the derived @xmath203 s for the disk are @xmath204 times larger than for the halo .
both effects are largely due to the relatively small number of stars in the disk solutions .
again we adopt disk-3 as the best solution .
since the disk velocity ellipsoid is quite `` round '' ( @xmath205 ) , a bias correction of @xmath206 mag ( figure 5 ) was subtracted from the solution value , giving @xmath207 } \rangle = -0.76.\ ] ] to see what would have happened had we been unable to separate the disk and halo rr lyraes , we performed a stat-@xmath18 solution ( last line of table 6 ) using all 213 stars .
this solution is comparable to `` group rr @xmath77 '' of hjbw ( 142 stars ) and to `` sample c@xmath208 '' ( 139 stars ) of srm .
interestingly , the absolute magnitude ( @xmath209 ) for our `` all stars '' solution is almost exactly the same as for the three halo solutions , but the kinematics are rather different .
for this mixture of disk and halo , @xmath132 , @xmath126 , and @xmath210 are smaller than for the pure halo , but @xmath14 is larger .
the velocity ellipsoid is much `` rounder '' ; @xmath211 , vs. 1.65 for halo-3 .
hjbw and srm found @xmath212 1.25 and 1.29 , respectively , for the comparable groups .
these results point out that a good disk / halo separation is necessary to get reliable kinematic results for the rr lyraes .
the stat-@xmath18 method is robust enough to produce solutions for mixed populations with distinctly different kinematics , successfully determining @xmath107 , but it derives kinematics not accurately representing either population . the fact that we have separate disk and halo solutions over a range of almost 1 dex in [ fe / h ] gives hope that we might obtain the slope of the @xmath9[fe / h ] relationship , a parameter of considerable astrophysical importance ( sec .
1 ) which previous stat-@xmath18 studies ( e.g. , hjbw , srm ) found difficult to determine . unfortunately ,
our disk solutions are not sufficiently precise to meaningfully constrain this slope .
figure 6 depicts this fact graphically ; the error bars on the disk solutions easily admit slopes between 0 and + 0.4 mag dex@xmath5 , the extreme values currently under debate . calculating the slope using the bias - corrected @xmath9 estimates for halo-3 and disk-3 from table 6
, we obtain @xmath213 = + 0.09 \pm 0.38\ mag\ dex}^{-1}.$ ] to pursue the @xmath9[fe / h ] slope further , we divided the halo-1 sample at [ fe / h ] = 1.55 , giving equal - sized metal - rich and metal - poor sub - groups .
we performed stat-@xmath18 solutions on each group ( halo-1r and halo-1p in table 6 , and the crosses in figure 6 ) .
these solutions show no indication of any slope within the halo .
again , we are unable to constrain the slope of the @xmath9[fe / h ] relation within meaningful limits . in a final effort to determine the @xmath9[fe / h ] slope , we attempted to incorporate this dependence directly into the stat-@xmath18 program by parameterizing the absolute magnitude term @xmath108 in eqn .
5 of hjbw with with two coefficients @xmath214 and @xmath215 , where @xmath216 - \langle [ fe / h ] } \rangle)\ + \ b.\ ] ] however , preliminary solutions indicate that this makes the maximum - likelihood solution ill - conditioned .
it may be possible to avoid this problem by incorporating the [ fe / h ] dependence into the distance scale correction @xmath104 rather than @xmath108 ; we are continuing to work on this problem , and any results will be reported in a future paper .
in sec . 1 , we mentioned the long history of efforts to determine @xmath9 .
csj provide an extensive review of much of this work , which we shall not repeat here .
table 7 presents the results of some of these efforts .
values produced by the stat-@xmath18 method are for `` typical '' rr lyraes in the sample .
these stars tend to be somewhat evolved off the zero age horizontal branch ( zahb ) .
some @xmath9 estimates using different methods refer to @xmath9 at the zahb .
we note these cases in table 7 , where we have corrected @xmath217 to @xmath9 using the eqn . 4 of csj .
] they are plotted as a function of [ fe / h ] in figure 7 to facilitate comparison with the results of our stat-@xmath18 solutions .
we refer the interested reader to the individual references in table 7 for more detailed discussions on these methods .
we begin by noting that our stat-@xmath18 solutions are in good agreement with the results of the two recent applications of the method ( @xmath218 , bh86 ; @xmath219 , srm ) .
this is not a complete surprise , given that the stars in those studies are all included in the present study , albeit with improved data .
as shown in table 7 and figure 7 , the agreement between the various stat-@xmath18 solutions becomes even better when small corrections are made to bring the previous results onto the system of reddenings and magnitudes used in this paper .
like us , neither of the previous groups was able to detect a meaningful trend in @xmath9 with [ fe / h ] due to the small sample sizes .
at the characteristic abundance of the halo , [ fe / h ] @xmath220 1.6 , the various results shown in figure 7 cover a range of 0.20.3 mag , and appear to separate into a brighter and a fainter group .
interestingly , whether a particular result is bright or faint does not seem to be a function of the method employed .
for example , buonanno _ et al . _
( 1990 ) found a bright zero - point by fitting 19 globulars to 5 subdwarfs , while csj found a faint zero - point by fitting a single cluster to the subdwarf ( of identical abundance ) with the best trigonometric parallax .
similarly , using the sandage period - shift effect , sandage ( 1990b ) obtained a bright zero - point and a steep slope . using data for a different set of field stars , adopting a different effective temperature relation , and employing a different mass - metallicity relation ,
csj obtained a moderate slope .
fernley ( 1993 ) used infra - red rather than optical observations of field and cluster rr lyraes in his period shift analysis , and found a bright zero - point but a moderate slope .
apparently , the current uncertainty in @xmath9 is dominated by differences in the details of the methods a particular author follows , and his or her choice of a particular data set or reddening correction .
it is very difficult to determine which of these assumptions are correct or incorrect at this level of detail . in deciding
which methods shown in figure 7 should be given the most weight , it is worth noting several strengths of the stat-@xmath18 method .
first , stat-@xmath18 is independent of other distance determinations .
by contrast , cluster main sequence fitting requires precise trigonometric parallaxes to the nearby subdwarfs .
calibrations of @xmath9 based on lmc distances determined by other methods are similarly complicated . for example , the cepheid calibration used by walker ( 1992 ) to obtain the lmc distance is based on main sequence fits of cepheid - bearing galactic open clusters to the pleiades , @xmath221 main sequence fits of the pleiades to local dwarfs with trigonometric parallaxes .
a second strength of the stat-@xmath18 method is that it relies on a simple , extremely well - tested model .
the kinematics of the galaxy are described by three mean velocities , three velocity dispersions , and the orientation of this velocity ellipsoid relative to the cardinal directions of the galaxy .
countless kinematic studies over the past half century have shown this to be a very complete description of local galactic kinematics . by comparison ,
the models or basic assumptions on which many of the other techniques rely are far more complex , and tend not to be so well - tested .
for example , the baade - wesselink method depends on model atmospheres to determine both the surface brightness constant , @xmath222 , and the observed - to - pulsation velocity correction , @xmath223 ( jones _ et al .
both the baade - wesselink and period shift methods rely on a color - index @xmath78 effective temperature relation , and different authors advocate different relations ( sandage 1990b , csj , fernley 1993 , sandage 1993 ) .
the period shift method also relies heavily on the assumed rr lyrae mass - metallicity relation , yet the mass estimates from double - mode rr lyraes are in serious conflict with those derived from the baade - wesselink method , and to a lesser extent with masses derived from hb theory ( fernley 1993 , yi _
meanwhile , @xmath9 estimates from hb theory are dependent on the color temperature relation , the evolutionary models ( including the treatment of convection ) , and especially on the assumed main sequence helium abundance , @xmath224 ( lee 1990 ) .
@xmath9 values derived from main sequence fits currently rely on theoretical isochrones to correct the colors and/or magnitudes of the clusters and/or field subdwarfs to a common metallicity ( buonanno _ et al .
_ 1990 , bolte & hogan 1995 ) .
the reader is referred to the papers noted above and in table 7 for detailed discussion of these topics .
our @xmath9 result for the halo is 0.06 mag brighter than the csj value at [ fe / h ] = 1.61 , and agrees better with this and other `` faint '' @xmath9 values than the `` bright '' results shown in fig .
7 , which are @xmath170.2 mag brighter . given this dichotomy , it is valid to ask whether there are any parameters we could change that would push our result to a brighter value .
the results of srm suggest that if we adopted the sturch ( 1966 ) reddening scale , based on the blanketing - corrected colors of rr lyraes at minimum light , we would obtain a result @xmath170.11 mag _ brighter _ than our current result .
however , srm argue that the burstein & heiles ( 1982 ) based redding scale is preferred . in sec .
2.4 , we noted that the various sources of rr lyrae photometry suffer small inconsistencies in photometric standardization at the 0.07 mag level .
had we used the un - transformed literature values , we would have obtained an @xmath9 for the halo about 0.07 mag _
fainter_. several other effects could alter our @xmath9 result by very small amounts . in sec .
4.6 , we showed that observations constrain @xmath105 to be between 0.0 and 0.1 . by adopting @xmath156
, we obtained the brightest @xmath9 consistent with this constraint ; adopting a smaller value of @xmath105 results in values of @xmath9 up to 0.03 mag _
fainter_. had we corrected for the small bias uncovered by our halo simulations ( sec .
4.8 ) , our result would have been @xmath170.04 mag _ brighter_. had we retained the 3 velocity dispersion covariance parameters in our model ( sec .
4.7 ) , our result would have been 0.02 mag _
brighter_. had we adopted the sample halo-2 rather than halo-3 , our result would have been @xmath170.01 mag _
fainter_. a final note in favor of our @xmath9 zero - point comes from the derived kinematics ( e.g. , sec .
if the brighter sandage ( 1993 ) @xmath9 is used , the velocity dispersions for the local thick disk and halo grow by 5 - 10% .
the dispersions derived from our stat-@xmath18 solutions and presented in table 6 are in good agreement with estimates of the velocity ellipsoids of these components as measured by other tracers ( in fact , the @xmath126 value for the halo is already larger than many estimates ) . enlarging them to match
the brighter @xmath9 degrades this agreement .
as discussed in sec .
1 , the @xmath9[fe / h ] relation , in particular its zero - point , is important in determining a number of quantities of interest to both galactic and extra - galactic astronomy . in this section
, we present the implications of our stat-@xmath18derived @xmath9 zero - point . in sec .
5.4 , we were unable to obtain a meaningful value for the slope of the @xmath9[fe / h ] relation . however , our zero - point for the halo rr lyraes is quite well determined . in the following discussion ,
we adopt our zero - point along with a slope @xmath225[fe / h ] = 0.15 mag dex@xmath5 , in agreement with hb theory ( lee 1990 ) , rgb theory ( fusi pecci _ et al . _ 1990 ) , baade - wesselink observations ( jones _ et al . _
1992 ) , and some analyses of the sandage period shift effect ( csj , fernley 1993 ) .
specifically , we adopt the relation @xmath9 = 0.15[fe / h ] + 0.95 mag .
walker & mack ( 1986 ) obtained the distance to the galactic center , @xmath226 , by finding the peak in the space density of rr lyraes as a function of distance along the line of sight through baade s window ( bw ) .
they recalibrated the photographic photometry of blanco ( 1984 ) to the johnson @xmath227 band using several ccd standard fields , corrected it for interstellar absorption , and converted it to the @xmath68 band using a @xmath228 period relation obtained from ngc 6171 .
they found @xmath229 if @xmath9 = + 0.60 mag .
we repeat their analysis here , comparing the results obtained using both their preferred value of @xmath9 = + 0.60 mag , and our preferred value of @xmath9 = 0.15[fe / h ] + 0.95 mag .
we employ the newer reddening estimates of @xmath230 = 0.50 for bw ( walker & terndrup 1991 ) and @xmath230 = 0.33 for ngc 6171 ( harris 1994 ) .
we also employ the walker & terndrup ( 1991 ) @xmath66@xmath67 metallicities , rather than the photometric ones used by walker & mack ( 1986 ) . like those authors
, we found that a small shift ( + 0.04 mag ) should be applied to make the fitted line in the @xmath231 period plane coincide with the bw rr lyraes listed in their table 7 , presumably to correct for slight inconsistencies in the adopted reddenings and/or metallicities of the cluster rr lyraes relative to those in bw .
figure 8 shows the space density of stars ( in arbitrary units ) as a function of distance through bw . using a variety of methods
, we find that the curve based on @xmath9 = 0.15[fe / h ] + 0.95 peaks at @xmath232 = 7.4 kpc ( squares ) , while that based on @xmath9 = + 0.60 peaks at @xmath232 = 8.1 kpc ( circles ) . after applying the small geometric correction of 1.03 discussed by walker & mack ( 1986 )
, we obtain @xmath233 and @xmath234 , respectively .
we note that the relation @xmath9 = 0.15[fe / h ] + 0.725 mag , based on the walker ( 1992 ) zero - point ( see table 7 ) produces results almost identical with those of the @xmath235 relation .
we also note that the widths of the density curves produced using these three @xmath9[fe / h ] relations are nearly identical , after they are corrected for the distance scale effects produced by the different distance zero - points .
this supports the statement by walker & terndrup ( 1991 ) that the abundance dispersion in bw is too small to meaningfully constrain the slope of the @xmath9[fe / h ] relation using this method .
the short distance to the galactic center based on our stat-@xmath18 zero - point is favored by the `` primary '' distance source , h@xmath236o maser proper motions ( @xmath237 kpc ) , quoted by reid ( 1993 ) .
reid lists a number of other @xmath226 estimates , and from them derives a `` best value '' of @xmath238 kpc . as noted by carney _
( 1995 ) , this value is in part determined using a `` bright '' rr lyrae calibration .
if we recompute the `` best value '' excluding the optical rr lyrae - dependent methods , we obtain @xmath239 kpc .
this value is further supported by carney _
( 1995 ) , who recently found @xmath240 kpc from the @xmath241-band photometry of 58 rr lyrae stars in bw , as calibrated using the @xmath242log(@xmath243 ) relation of csj . the age of a globular cluster can be determined by comparing the absolute magnitude of its main - sequence turnoff ( msto ) with the mstos of theoretical isochrones .
the difference in apparent magnitude between a cluster s rr lyraes and its msto , together with an adopted value for @xmath9 , can be used to find the absolute magnitude of the cluster s msto .
it is instructive to see the difference between the ages derived using our @xmath9 zero - point and those using the brighter zero - point of walker ( 1992 ) . for both cases ,
we adopt a metallicity dependence of 0.15 mag dex@xmath5 , so the relations are @xmath244 [ fe / h ] + 0.95 , and @xmath244 [ fe / h ] + 0.725 , respectively .
the zero - point of the @xmath9[fe / h ] relation primarily determines the mean age of the globular cluster system , while the slope of the relation determines the age distribution .
thus our comparison focuses on the mean ages of the halo globular cluster system under the two @xmath9 zero - points .
brian chaboyer has kindly computed ages for the 39 `` older '' clusters listed in table 3 of chaboyer , demarque & sarajedini ( 1996b ) using both of these @xmath9[fe / h ] relations .
the ages were derived using the opal equation of state isochrones of chaboyer & kim ( 1995 ) .
these isochrones include the effects of diffusion and the latest available equation of state ( rogers 1994 ) , and employ modern helium and @xmath245-element abundances .
the weighted mean age of the 39 clusters is @xmath246 gyr using our zero - point , and @xmath247 gyr using the walker ( 1992 ) zero - point ( standard errors of the mean ) .
we account for the uncertainties in the @xmath9 zero - points by computing the mean ages from @xmath9[fe / h ] relations based on zero - points of @xmath248 ( again , thanks to b. chaboyer ) . for our stat-@xmath18
@xmath9 zero - point , the mean age including formal errors is @xmath249 gyr , compared with @xmath250 gyr using the walker ( 1992 ) zero - point .
these values are @xmath1714% smaller than the ages computed without the improved treatments of diffusion and the equation of state ( chaboyer _ et al .
_ 1996b ) .
clearly , the @xmath9 zero - point derived from the present stat-@xmath18 analysis supports an older age for the globular cluster system .
the ages of the _ oldest _ globular clusters place a lower limit on the age of the universe .
( 1996a ) define a group of 17 clusters which they suspect represents the oldest galactic globulars . using the method described above
, their weighted mean age is @xmath251 gyr using our stat-@xmath18 @xmath9 zero - point , and @xmath252 gyr using walker s zero - point .
a word of caution is warranted here .
the median [ fe / h ] of this group is @xmath170.2 dex lower than that of the field rr lyraes used to determine the @xmath9 zero - point , so if the @xmath9[fe / h ] slope is 0.30 ( 0.0 ) mag dex@xmath5 , the mean derived ages are @xmath253 gyr smaller ( larger ) than those quoted .
walker ( 1992 ) lists the de - reddened @xmath70 magnitudes of the rr lyrae stars in seven large magellanic cloud ( lmc ) globular clusters .
using an lmc distance modulus of 18.50 @xmath75 0.10 , based on lmc cepheid observations and an abundance - corrected galactic cepheid calibration , he obtained @xmath254 at [ fe / h ] = 1.9 , consistent with the `` brighter '' @xmath9 values shown in fig .
7 . this distance modulus implies an lmc distance of @xmath255 kpc .
using walker s @xmath70 magnitudes , and adopting @xmath256[fe / h ] + 0.95 , based on our new stat-@xmath18 zero - point , we find the lmc distance modulus to be 18.28 @xmath75 0.13 , equivalent to a distance of 45 @xmath75 3 kpc . given the complexities of deriving the cepheid period - luminosity relation ( see sec .
6 , also walker ( 1992 ) and references therein ) , it seems worthwhile to explore the consequences of applying a zero - point offset to the cepheid period - luminosity relation to make it match the shorter lmc distance based on our stat-@xmath18 results .
a particularly interesting consequence involves the measurement of the hubble constant , @xmath13 .
many extra - galactic distance indicators are calibrated to , or are consistent with , an lmc distance modulus of 18.50 mag .
if the distances determined using these indicators are recalibrated to agree with our smaller lmc distance modulus , the value of @xmath13 derived from them increases by 10% .
for example , the recent cepheid - based distance of @xmath257 mpc to m100 yielded @xmath258 km s@xmath5 mpc@xmath5 ( ferrarese _ et al .
if we reduce the m100 distance modulus by 0.22 mag to bring the cepheid period - luminosity relation into agreement with our lmc distance , we obtain @xmath259 mpc and @xmath260 km s@xmath5 mpc@xmath5 .
freedman _ et al . _
( 1994 ; their fig .
3 ) have already shown that the expansion age of the universe implied by @xmath261 km s@xmath5 mpc@xmath5 in the framework of the einstein de sitter cosmological model is in conflict with the observed ages of globular clusters .
our stat-@xmath18 @xmath9 zero - point indicates a larger value of @xmath13 ( shorter expansion time ) and older globular clusters , thus increasing the disagreement between these two important observables .
the disagreement persists even at lower values of the density parameter ( @xmath262 ) .
however , other recent measurements of @xmath13 obtain lower values which are in better agreement with our cluster ages .
for example , branch _ et al . _
( 1996 ) obtained @xmath263 km s@xmath5 mpc@xmath5 from type ia supernovae .
additional possible routes to reconciling observations of @xmath13 and globular cluster ages include accepting a non - zero cosmological constant ( carroll & press 1992 ) and further refinements to stellar evolution theory which result in younger cluster ages ( e.g. , mazzitelli _ et al .
we have assembled high - quality data on 213 nearby rr lyrae variables .
these data include new absolute proper motions from the lick northern proper motion program ( klemola _ et al .
_ 1993 ) and abundances and radial velocities from layden ( 1994 ) . based on an _ a priori _ kinematic study
, we defined three ways to separate the stars into thick disk and halo sub - populations . statistical parallax solutions for these sub - samples yielded the absolute magnitude and kinematics of the rr lyraes in the samples .
we note that our @xmath9 values correspond to the absolute magnitude of typically - evolved rr lyrae stars , not to that of the zero age horizontal branch . for the halo population ,
the solutions produced a well determined absolute magnitude , @xmath0 at @xmath1[fe / h]@xmath2 = 1.61 .
the derived kinematics , @xmath264 and @xmath265 , are in good agreement with previous estimates of the halo rr lyrae kinematics , and with the kinematics of other stellar tracers of the halo . for the thick disk population , the results of the three definitions scatter somewhat , and the uncertainties are larger .
our best estimate for the thick disk was @xmath3 at @xmath1[fe / h]@xmath2 = 0.76 , after correction for the bias mentioned below .
the derived kinematics , @xmath266 and @xmath267 , again are in good agreement with previous estimates of the thick disk kinematics based on rr lyraes and on other tracers .
the large uncertainty in the disk solution prevented us from deriving a meaningful slope for the @xmath9[fe / h ] relation .
an attempt to measure the slope by sub - dividing the halo sample into two metallicity bins also failed to meaningfully determine the slope .
monte carlo tests using simulated data showed that our stat-@xmath18 code accurately returns the true kinematic and @xmath107 values of the input data .
they also revealed the possibility that the internal errors returned by the hjbw stat-@xmath18 algorithm may be overestimates .
for the halo , the true error of @xmath9 may be @xmath170.08 mag rather than 0.12 mag .
the kinematic results for both the disk and halo may also be more precise than the errors cited above .
however , it is outside the scope of this paper to determine which set of error values is correct .
so , for the present , we have conservatively adopted the larger values .
the simulations also enabled us to evaluate the effects of other factors on the solutions .
all were negligible , except for a small bias towards `` short '' distance scales that is observed when the @xmath80 and @xmath68 velocity dispersions are of comparable size .
we determined corrections based on the simulations : 0.15 mag for the disk and 0.04 mag for the halo solutions .
the former correction was applied to the real - data solutions , while the latter was deemed too small to be of practical value .
we discussed the effects of systematic errors .
the main systematic uncertainty is in the adopted reddening scale .
our scale ( burstein & heiles 1982 ) makes our @xmath9 value @xmath170.11 mag fainter than it would have been using the sturch ( 1966 ) reddenings , but srm argue that the former scale is preferred .
our adopted photometric scale makes our quoted @xmath9 value as bright as possible ( @xmath170.07 mag ) .
several other minor systematics , including the halo bias correction noted above , were also considered .
it is unlikely that these systematic errors alone can account for the difference between our @xmath9 value and the brighter ( @xmath170.2 mag ) values found by some authors ( e.g. , buonanno _ et al .
_ 1990 ; sandage 1990b , 1993 ; walker 1992 ) . unlike the methods used by many authors ,
the stat-@xmath18 method is independent of other distance calibrations .
it also depends on a relatively simple and well - tested model ( galactic kinematics ) in comparison to the other methods , which employ model atmospheres , stellar evolution theory , empirical color temperature relations , rr lyrae mass determinations , etc .
we investigated the implications of our halo @xmath9 by using an @xmath9[fe / h ] relation based on an adopted slope of @xmath15[fe / h ] = 0.15 mag dex@xmath5 in combination with our halo zero - point .
we found the distance to the galactic center to be @xmath268 kpc based on observations of the rr lyraes in baade s window , in good agreement with many other estimates of @xmath226 . using the `` brighter '' @xmath9 values
, @xmath226 is 10% larger .
following chaboyer _
( 1996b ) , we found the mean age of the 17 oldest galactic globular clusters to be @xmath269 gyr , 3.4 gyr older than the mean age obtained using the brighter rr lyrae zero - point ; this places an important lower limit on the age of the universe .
we found the distance modulus of the large magellanic cloud to be @xmath12 mag .
any estimates of the hubble constant , @xmath13 , which are based on an lmc distance modulus of 18.50 mag ( e.g. , the cepheid study of ferrarese _ et al .
_ 1996 ) increase by 10% if their distance scales are recalibrated to match our lmc distance .
this increase implies a younger age for the universe , in conflict with the older globular cluster ages derived from our @xmath9 value .
the conflict is lessened or eliminated if the true value of @xmath13 is low , if the cosmological constant is non - zero , or if further refinements to stellar evolution theory result in younger cluster ages .
the authors acknowledge valuable discussions with drs .
brian chaboyer , jan lub , robert schommer , and alistair walker .
the comments of an anonymous referee improved several sections of the paper .
acl acknowledges financial support from the natural sciences and engineering research council of canada , through research grants to w.e .
harris and d.l .
welch , and from cerro tololo inter - american observatory .
the lick northern proper motion program is supported by national science foundation grant ast 92 - 18084 .
slh acknowledges the support of nsf young investigator ( nyi ) grant number ast94 - 57455 .
cjh performed part of this research as an reu student at michigan state university .
ajhar , e.a . ,
grillmair , c.j . , lauer , t.r . ,
baum , w.a . ,
faber , s.m . , holtzman , j.a .
, lynds , c.r . &
oneil , e.j.jr .
1996 , , 111 , 1110 barnes iii , t.g . , & hawley , s.l .
1986 , , 307 , l9 ( bh86 ) beers , t.c .
1996 , formation of the galactic halo ... inside and out , asp conf .
morrison & a. sarajedini , ( asp : san francisco ) , 130 blanco , b.m .
1984 , , 89 , 1836 blanco , v.m .
1992 , , 104 , 734 bolte , m. & hogan 1995 , nature , 376 , 399 bookmyer , b.b . ,
fitch , w.s . ,
lee , t.a . ,
wisniewski , w.z . & johnson , h.l .
1977 , rev .
astrofis . , 2 , 235 branch , d. , romanishin , w. , & baron , e. 1996 , , 465 , 73 buonanno , r. , corsi , c.e . & fusi pecci , f. 1989 , , 216 , 80 buonanno , r. , cacciari , c. , corsi , c.e .
& fusi pecci , f. 1990 , , 230 , 315 burstein , d. , & heiles , c. 1982 , , 87 , 1165 carney , b.w . , storm , j. & jones , r.v .
1992 , , 386 , 663 ( csj ) carney , b.w . ,
fulbright , j.p . ,
terndrup , d.m . ,
suntzeff , n.b . , & walker , a.r .
1995 , , 110 , 1674 carrol , s.m . & press , w.h .
1992 , , 30 , 499 casertano , s. , ratnatunga , k.u . , & bahcall , j.n .
1990 , , 357 , 435 chaboyer , b. 1995 , , 444 , l9 chaboyer , b. & kim , y .- c .
1995 , , 454 , 767 chaboyer , b. , demarque , p. , kerman , p.j . & krauss , l.m . 1996a , science , 271 , 957 chaboyer , b. , demarque , p. & sarajedini , a. 1996b , , 459 , 558 clube , s.v.m .
& dawe , j.a .
1980a , , 190 , 575 clube , s.v.m . & dawe , j.a .
1980b , , 190 , 591 ( cd80 ) croswell , k. , latham , d.w . & carney , b.w .
1987 , , 93 , 1445 eggen , o.j . , lynden - bell , d. & sandage , a. 1962 , , 136 , 748 fernley , j. 1993 , , 268 , 591 ferrarese , l. , _ et al .
_ 1996 , , 464 , 568 fitch , w.s . ,
wisniewski , w.z . & johnson , h.l .
1966 , comm .
lunar planet . lab .
71 , 3 fitzgerald , m.p .
1968 , , 73 , 983 fitzgerald , m.p .
1987 , , 229 , 227 freedman , w.l . , _
_ 1994 , nature , 371 , 757 fusi pecci , f. , ferraro , f.r . ,
crocker , d.a . ,
rood , r.t .
& buonanno , r. 1990 , , 238 , 95 gould , a. 1995 , , 452 , 189 harris , w.e .
1994 http://www.physics.mcmaster.ca/globular.html hanson , r.b .
1993 , lick northern proper motion program : npm1 catalog documentation for the computer - readable version , national space science data center document no .
nssdc / wdc - a - r&s 93 - 41 hanson , r.b .
, & klemola , a.r .
1994 , lick northern proper motion program : npm1 catalog cross - identifications and appendices documentation for the computer - readable version , national space science data center document no .
nssdc / wdc - a - r&s 94 - 14 hawley , s.l . ,
jefferys , w.h . ,
barnes iii , t.g .
& wan , l. 1986 , , 302 , 626 ( hjbw ) hemenway , m.k .
1975 , , 80 , 199 johnson , d.r.h . & soderblom , d.r .
1987 , , 93 , 864 jones , r.v . ,
carney , b.w .
, & latham , d.w .
1988 , , 332 , 206 jones , r.v . ,
carney , b.w .
, storm , j. & latham , d.w .
1992 , , 386 , 646 joy , a.h . 1950 , , 62 , 60 jurcsik , j. & kovcs , g. 1996 , , in press kerr , f.j . & lynden - bell , d. 1986 , , 221 , 1023 kholopov , p.n .
1985 , general catalog of variable stars , 4th ed . , ( nauka , moscow ) klemola , a.r . ,
hanson , r.b . & jones , b.f .
1993 , lick northern proper motion program : npm1 catalog , national space science data center astronomical data center catalog no .
a1199 klemola , a.r . ,
hanson , r.b . &
jones , b.f .
1994a , lick northern proper motion program : npm1 catalog cross - identifications and appendices , national space science data center astronomical data center catalog no .
a1214 klemola , a.r . , hanson , r.b .
& jones , b.f .
1994b , lick npm program : npm1 catalog and its applications , in galactic and solar system optical astrometry , eds .
morrison and g.f .
gilmore , cambridge , cambridge university press , pp .
20 - 25 klemola , a.r . ,
jones , b.f . & hanson , r.b .
1987 , , 94 , 501 lambert , d.l . , heath , j.e .
, lemke , m. , & drake , j. 1995 , , 104 , 183 landolt , a.u .
1992 , , 104 , 340 layden , a.c .
1994 , , 108 , 1016 ( l94 ) layden , a.c .
1995 , , 110 , 2288 ( l95 ) layden , a.c . , hawley , s.l . &
hanson , r.b .
1994 , , 184 , 3202 layden , a.c . , 1996 , in preparation lee , y .- w . ,
demarque , p. & zinn , r. 1990 , , 350 , 155 lee , y .- w .
1990 , , 363 , 159 lub , j. , 1977 , ph.d .
dissertation , leiden university lutz , t.e . & hanson , r.b .
1992 , asp conf .
series , 25 , 257 majewski , s.r .
1993 , , 31 , 575 mazzitelli , i. , dantona , f. & caloi , v. 1995 , , 302 , 382 mihalas , d. & binney , j. 1981 , galactic astronomy , second edition , ( freedman : san francisco ) morrison , h.l . , flynn , c. & freeman , k.c .
1990 , , 100 , 1191 nissen , p.e . & schuster , w.j .
1991 , , 251 , 457 murray , c.a . 1983 ,
vectorial astrometry , bristol , adam hilger ltd .
297 - 302 pel , j.w . , 1976 ,
a&a supp , 24 , 413 pierce , m.j . ,
welch , d.l . ,
mcclure , r.d .
, van den bergh , s. , racine , r. , & stetson , p.b .
1994 , nature , 371 , 385 preston , g.w .
1959 , , 130 , 507 reid , m.j .
1993 , , 31 , 345 rogers , f.j .
1994 , in the equation of state in astropyhsics , iau coll .
g. chabrier and e. schatzman ( cambridge : cambridge university press ) , 16 sandage , a. 1990a , , 350 , 603 sandage , a. 1990b , , 350 , 631 sandage , a. 1993 , , 106 , 703 schmidt , e.g. 1991 , , 102 , 1766 schmidt , e.g. , chab , j.r . , reiswig , d.e .
1995 , , 109 , 1239 searle , l. & zinn , r. 1978 , , 225 , 357 strugnell , p. , reid , n. & murray , c.a .
1986 , , 220 , 413 ( srm ) sturch , c.r .
( 1966 ) , , 143 , 774 walker , a.r . 1992 , , 390 , l81 walker , a.r .
& mack , p. 1986
, , 220 , 69 walker , a.r . &
terndrup , d.m .
1991 , , 378 , 119 wan , l. , he , m .- f . , zhu , g .-
l . , & li , z .- y . 1979 , chinese astronomy , 3 , 296 wan , l. , mao , y .- q .
& ji , d .- s . 1980 , ann .
shanghai obs .
, no . 2 , 1 ( wmj ) yi , s. , lee , y .- w .
& demarque , p. 1993
, , 411 , l25 zinn , r.j .
1993 , the globular cluster galaxy connection , asp conf .
smith & j.p .
brodie , ( asp : san francisco ) , 38 zinn , r. , & west , m.j .
1984 , , 55 , 45 sw and & + 29.0017 & 115.72 & 33.09 & 0.53 & 2.49 & 21 & 2 & 0.38 & 9.68 & 0.14 & 1121 & 111 xx and & + 38.0072 & 128.45 & 23.64 & 3.53 & 3.56 & 0 & 5 & 2.01 & 10.63 & 0.13 & 1111 & 000 xy and & + 33.0068 & 131.22 & 28.23 & 1.34 & 0.28 & 64 & 53 & 0.92 & 13.63 & 0.15 & 1161 & 000 zz and & + 26.0036 & 122.42 & 35.85 & 2.50 & 1.80 & 13 & 53 & 1.58 & 13.01 & 0.12 & 1141 & 000 ci and & + 43.0105 & 134.93 & 17.62 & 0.15 & 0.38 & 99 & 30 & 0.83 & 12.15 & 0.26 & 1141 & 111 dr and & + 33.0052 & 126.16 & 28.57 & 3.00 & 1.38 & 81 & 30 & 1.48 & 12.34 & 0.09 & 1141 & 000 wy ant & & 266.93 & 22.08 & 2.56 & 4.81 & 211 & 24 & 1.66 & 10.83 & 0.18 & 2111 & 000 sw aqr & 00.1628 & 51.31 & 31.47 & 4.67 & 6.08 & 42 & 8 & 1.24 & 11.14 & 0.22 & 1111 & 000 sx aqr & + 03.1345 & 57.90 & 34.00 & 3.87 & 5.14 & 166 & 7 & 1.83 & 11.70 & 0.11 & 1111 & 000 tz aqr & 05.1879 & 53.25 & 44.33 & 0.96 & 1.77 & 35 & 12 & 1.24 & 12.01 & 0.10 & 1121 & 000 yz aqr & 11.1939 & 48.93 & 49.76 & 1.25 & 0.21 & 150 & 14 & 1.55 & 12.65 & 0.08 & 1161 & 000 aa aqr & 10.2266 & 54.63 & 53.83 & 2.11 & 0.30 & 20 & 14 & 2.09 & 12.37 & 0.15 & 1161 & 000 bn aqr & 07.2815 & 56.22 & 50.73 & 0.96 & 3.26 & 182 & 30 & 1.33 & 12.52 & 0.10 & 1161 & 000 bo aqr & 12.2527 & 55.40 & 58.83 & 0.82 & 1.21 & 24 & 13 & 1.80 & 12.11 & 0.07 & 1111 & 000 br aqr & 09.2377 & 75.48 & 65.24 & 0.50 & 0.02 & 29 & 10 & 0.84 & 11.40 & 0.04 & 1121 & 111 bt aqr & 05.1748 & 42.94 & 30.60 & 0.30 & 1.06 & 52 & 11 & 0.29 & 12.34 & 0.12 & 1121 & 111 cp aqr & 01.1098 & 48.74 & 31.34 & 1.45 & 1.73 & 29 & 21 & 0.90 & 11.71 & 0.11 & 1131 & 111 dn aqr & & 35.76 & 69.06 & 4.60 & 1.60 & 214 & 8 & 1.63 & 11.18 & 0.02 & 2131 & 000 aa aql & 03.1592 & 43.08 & 24.99 & 1.80 & 1.54 & 32 & 4 & 0.58 & 11.74 & 0.21 & 1121 & 111 v341 aql & & 45.62 & 22.04 & 3.48 & 2.40 & 81 & 4 & 1.37 & 10.81 & 0.31 & 2121 & 000 x ari & + 10.0299 & 169.08 & 39.84 & 5.97 & 9.24 & 35 & 3 & 2.40 & 9.51 & 0.50 & 1111 & 000 tz aur & + 40.0228 & 176.79 & 20.92 & 0.30 & 1.35 & 46 & 6 & 0.80 & 11.86 & 0.18 & 1121 & 111 rs boo & + 31.0704 & 50.84 & 67.35 & 0.22 & 1.17 & 9 & 2 & 0.32 & 10.36 & 0.00 & 1121 & 111 st boo & + 35.0720 & 57.39 & 55.22 & 1.54 & 1.67 & 13 & 4 & 1.86 & 10.98 & 0.04 & 1121 & 000 sv boo & + 39.0683 & 68.75 & 65.51 & 0.17 & 2.28 & 131 & 22 & 1.55 & 13.12 & 0.00 & 1121 & 000 sw boo & + 36.0643 & 62.52 & 67.74 & 4.58 & 0.13 & 18 & 18 & 1.12 & 12.34 & 0.00 & 1111 & 000 sz boo & + 28.0840 & 41.93 & 65.50 & 0.77 & 0.85 & 38 & 21 & 1.68 & 12.60 & 0.01 & 1111 & 000 tw boo & + 41.0778 & 71.06 & 62.85 & 0.34 & 5.86 & 99 & 4 & 1.41 & 11.20 & 0.01 & 1121 & 000 uu boo & + 35.0710 & 56.50 & 58.01 & 0.86 & 3.38 & 10 & 28 & 1.92 & 12.22 & 0.01 & 1121 & 000 uy boo & + 13.0991 & 354.24 & 68.81 & 0.28 & 4.51 & 144 & 3 & 2.49 & 10.80 & 0.00 & 1121 & 000 rz cam & + 67.0052 & 147.98 & 23.17 & 0.75 & 0.57 & 266 & 26 & 1.01 & 12.73 & 0.18 & 1121 & 000 rw cnc & + 29.0370 & 197.49 & 43.53 & 0.79 & 4.17 & 85 & 7 & 1.52 & 11.85 & 0.03 & 1161 & 000 ss cnc & + 23.0300 & 198.94 & 26.28 & 0.77 & 1.72 & 27 & 16 & 0.07 & 12.16 & 0.09 & 1121 & 111 tt cnc & + 13.0361 & 212.10 & 28.38 & 4.60 & 4.00 & 49 & 5 & 1.58 & 11.24 & 0.12 & 1111 & 000 an cnc & + 15.0560 & 212.10 & 35.03 & 0.51 & 2.53 & 16 & 14 & 1.45 & 13.16 & 0.06 & 1141 & 000 aq cnc & + 12.0610 & 218.04 & 38.10 & 1.15 & 3.39 & 390 & 20 & 1.53 & 12.00 & 0.04 & 1141 & 000 as cnc & + 25.0305 & 197.89 & 31.23 & 2.79 & 0.83 & 258 & 26 & 1.89 & 12.50 & 0.05 & 1141 & 000 w cvn & + 38.0637 & 71.82 & 70.96 & 1.71 & 0.89 & 18 & 21 & 1.21 & 10.52 & 0.00 & 1111 & 001 z cvn & + 44.0884 & 124.00 & 73.35 & 0.79 & 3.97 & 14 & 10 & 1.98 & 11.93 & 0.00 & 1121 & 000
rr cvn & + 34.0596 & 154.05 & 81.09 & 1.58 & 3.17 & 5 & 21 & 1.08 & 12.55 & 0.01 & 1111 & 000 ru cvn & + 31.0668 & 53.96 & 74.51 & 2.97 & 0.22 & 27 & 21 & 1.37 & 11.96 & 0.00 & 1121 & 000
rx cvn & + 41.0711 & 87.08 & 71.53 & 0.20 & 0.10 & 158 & 28 & 1.31 & 12.57 & 0.00 & 1121 & 000
rz cvn & + 32.1089 & 61.59 & 77.15 & 5.84 & 0.36 & 12 & 7 & 1.92 & 11.42 & 0.00 & 1121 & 000
ss cvn & + 40.0608 & 83.85 & 72.63 & 1.24 & 5.10 & 15 & 22 & 1.52 & 11.89 & 0.00 & 1121 & 000
sv cvn & + 37.0978 & 139.98 & 79.40 & 0.12 & 2.55 & 29 & 28 & 2.20 & 12.59 & 0.00 & 1121 & 000
sw cvn & + 37.0987 & 134.84 & 79.80 & 0.96 & 1.98 & 18 & 21 & 1.53 & 12.74 & 0.00 & 1121 & 000 uz cvn & + 40.0532 & 139.53 & 75.93 & 0.06 & 2.32 & 38 & 26 & 2.34 & 12.02 & 0.00 & 1141 & 000 al cmi & + 05.0339 & 214.43 & 15.35 & 1.21 & 0.52 & 46 & 21 & 0.85 & 12.01 & 0.09 & 1151 & 111 rv cap & 15.2350 & 33.13 & 35.54 & 2.42 & 10.57 & 106 & 7 & 1.72 & 10.92 & 0.11 & 1111 & 000 iu car & & 269.59 & 22.95 & 1.30 & 0.60 & 328 & 18 & 1.85 & 11.91 & 0.39 & 2111 & 000 v499 cen & & 315.10 & 18.12 & 2.10 & 0.25 & 323 & 24 & 1.56 & 11.05 & 0.18 & 2111 & 000 dx cep & + 83.0238 & 119.50 & 21.95 & 1.86 & 0.91 & 6 & 30 & 1.83 & 12.67 & 0.35 & 1161 & 000 rr cet & + 01.0121 & 143.54 & 59.89 & 0.05 & 3.88 & 75 & 1 & 1.52 & 9.75 & 0.02 & 1111 & 001 ru cet & 16.0129 & 134.27 & 78.63 & 2.15 & 1.00 & 57 & 8 & 1.60 & 11.60 & 0.01 & 1111 & 000 rv cet & 11.0305 & 177.32 & 64.40 & 2.58 & 1.83 & 93 & 7 & 1.32 & 10.76 & 0.02 & 1111 & 000 rx cet & 15.0060 & 102.48 & 77.64 & 2.89 & 6.25 & 58 & 7 & 1.46 & 11.36 & 0.03 & 1111 & 000 rz cet & 08.0299 & 178.21 & 60.34 & 2.59 & 1.88 & 10 & 6 & 1.50 & 11.84 & 0.03 & 1121 & 000 xz cet & 16.0267 & 182.40 & 70.75 & 4.14 & 0.13 & 167 & 10 & 2.27 & 9.49 & 0.00 & 1161 & 000 ry col & & 246.47 & 35.05 & 3.60 & 1.80 & 482 & 15 & 1.11 & 10.86 & 0.01 & 2131 & 000 s com & + 27.0979 & 213.16 & 85.84 & 2.22 & 2.01 & 55 & 4 & 2.00 & 11.55 & 0.02 & 1111 & 000 v com & + 27.0931 & 208.67 & 80.85 & 0.84 & 0.47 & 23 & 28 & 1.75 & 13.16 & 0.02 & 1121 & 000 ry com & + 23.0619 & 342.56 & 85.06 & 0.61 & 1.78 & 31 & 8 & 1.65 & 12.30 & 0.04 & 1121 & 000 tv crb & + 27.1359 & 41.33 & 56.51 & 0.31 & 1.01 & 157 & 48 & 2.33 & 11.61 & 0.07 & 1161 & 000 w crt & 17.1378 & 276.00 & 40.47 & 2.38 & 1.48 & 65 & 13 & 0.50 & 11.51 & 0.09 & 1121 & 111 x crt & 10.1333 & 278.87 & 49.49 & 0.13 & 3.55 & 79 & 4 & 1.75 & 11.48 & 0.00 & 1111 & 000 xz cyg & & 88.21 & 16.98 & 8.43 & 2.15 & 119 & 14 & 1.52 & 9.72 & 0.33 & 2111 & 000 dm cyg & + 31.0966 & 79.46 & 12.41 & 1.34 & 0.72 & 12 & 23 & 0.14 & 11.49 & 0.69 & 1121 & 111 dx del & + 12.1761 & 58.47 & 18.84 & 1.41 & 1.17 & 45 & 3 & 0.56 & 9.92 & 0.32 & 1121 & 111 rw dra & + 57.0740 & 87.39 & 40.60 & 0.23 & 0.82 & 108 & 22 & 1.40 & 11.57 & 0.00 & 1121 & 001 su dra & + 67.0274 & 133.44 & 48.27 & 4.95 & 7.26 & 167 & 1 & 1.74 & 9.78 & 0.00 & 1111 & 000 sw dra & + 69.0237 & 127.27 & 47.33 & 3.22 & 0.93 & 30 & 1 & 1.24 & 10.49 & 0.04 & 1121 & 001 wy dra & + 80.0305 & 113.08 & 25.14 & 1.11 & 0.88 & 6 & 30 & 1.66 & 12.67 & 0.24 & 1161 & 000 xz dra & + 64.0579 & 95.65 & 22.50 & 0.56 & 0.71 & 30 & 2 & 0.87 & 10.18 & 0.22 & 1121 & 111 ae dra & + 55.0697 & 84.35 & 25.41 & 0.83 & 0.24 & 243 & 30 & 1.54 & 12.65 & 0.14 & 1161 & 000 bc dra & + 76.0341 & 107.95 & 28.48 & 2.18 & 3.30 & 161 & 26 & 2.00 & 11.60 & 0.18 & 1161 & 000 bd dra & + 77.0411 & 108.63 & 28.25 & 2.88 & 0.07 & 253 & 30 & 1.74 & 12.69 & 0.10 & 1161 & 000 bt dra & + 60.0453 & 99.41 & 51.21 & 0.13 & 3.60 & 156 & 30 & 1.55 & 11.94 & 0.00 & 1161 & 000
rx eri & 15.0672 & 214.26 & 33.88 & 1.56 & 0.70 & 66 & 1 & 1.30 & 9.68 & 0.08 & 1111 & 000 sv eri & 11.0414 & 194.26 & 53.47 & 1.45 & 3.95 & 12 & 9 & 2.04 & 9.95 & 0.19 & 1111 & 000 xy eri & 13.0580 & 207.42 & 41.69 & 1.13 & 0.68 & 221 & 11 & 2.08 & 13.02 & 0.08 & 1161 & 000 bb eri & 19.0632 & 218.81 & 34.36 & 3.45 & 0.91 & 235 & 11 & 1.51 & 11.46 & 0.03 & 1121 & 000 bk eri & 01.0185 & 175.80 & 51.70 & 3.24 & 1.98 & 141 & 10 & 1.64 & 12.67 & 0.10 & 1161 & 000 ss for & & 216.42 & 72.99 & 4.30 & 7.25 & 112 & 1 & 1.35 & 10.10 & 0.00 & 2121 & 000 sw for & & 243.27 & 60.75 & 1.25 & 0.10 & 174 & 18 & 1.95 & 12.34 & 0.00 & 2111 & 000 rr gem & & 187.44 & 19.52 & 0.35 & 0.24 & 64 & 1 & 0.35 & 11.34 & 0.21 & 2121 & 111 sz gem & + 19.0314 & 201.85 & 22.08 & 1.17 & 3.40 & 307 & 11 & 1.81 & 11.66 & 0.08 & 1121 & 000 tw her & + 30.0973 & 55.87 & 24.80 & 0.32 & 0.74 & 4 & 16 & 0.67 & 11.23 & 0.17 & 1121 & 111 vz her & + 36.0784 & 59.59 & 34.59 & 2.44 & 1.95 & 115 & 4 & 1.03 & 11.44 & 0.12 & 1121 & 000 af her & + 41.0868 & 65.15 & 41.64 & 1.62 & 0.72 & 268 & 9 & 1.94 & 12.82 & 0.00 & 1121 & 000 ag her & + 40.0809 & 64.49 & 41.46 & 2.15 & 1.91 & 103 & 21 & 2.01 & 12.66 & 0.00 & 1121 & 000 cw her & + 35.0792 & 58.01 & 38.99 & 1.65 & 1.31 & 285 & 30 & 2.09 & 12.47 & 0.05 & 1141 & 000 dl her & + 14.1435 & 36.28 & 26.60 & 1.13 & 0.13 & 61 & 14 & 1.32 & 12.37 & 0.35 & 1141 & 101 gy her & + 37.1381 & 60.70 & 41.71 & 0.29 & 1.13 & 157 & 53 & 1.92 & 12.57 & 0.01 & 1141 & 000 v394 her & + 17.1675 & 40.01 & 27.39 & 0.67 & 0.67 & 74 & 10 & 1.48 & 12.87 & 0.21 & 1161 & 000 sv hya & & 297.08 & 36.59 & 5.51 & 0.81 & 100 & 8 & 1.70 & 10.51 & 0.34 & 2111 & 000 sz hya & 09.0958 & 239.77 & 25.93 & 0.66 & 3.85 & 140 & 9 & 1.75 & 11.23 & 0.05 & 1121 & 000 uu hya & + 04.0500 & 230.41 & 38.18 & 2.51 & 1.12 & 295 & 14 & 1.65 & 12.27 & 0.05 & 1121 & 000 wz hya & 12.1166 & 254.26 & 34.41 & 0.14 & 1.49 & 304 & 8 & 1.30 & 10.82 & 0.26 & 1121 & 000 xx hya & 15.1036 & 244.59 & 21.35 & 1.86 & 2.92 & 32 & 20 & 1.33 & 11.89 & 0.15 & 1121 & 000 dd hya & + 02.0597 & 219.85 & 19.30 & 0.70 & 1.19 & 153 & 25 & 1.00 & 12.18 & 0.06 & 1151 & 011 dg hya & 05.0893 & 233.78 & 24.95 & 1.08 & 2.52 & 164 & 18 & 1.42 & 12.14 & 0.06 & 1121 & 000 dh hya & 09.0936 & 238.03 & 22.96 & 2.37 & 0.67 & 355 & 8 & 1.55 & 12.13 & 0.05 & 1111 & 000 et hya & 08.0736 & 233.50 & 18.31 & 1.66 & 0.73 & 320 & 20 & 1.69 & 12.06 & 0.08 & 1151 & 000 fy hya & & 318.76 & 31.37 & 4.12 & 0.04 & 82 & 24 & 2.33 & 12.46 & 0.15 & 2111 & 000 gl hya & + 02.0649 & 223.71 & 25.46 & 1.34 & 0.90 & 223 & 21 & 1.45 & 12.95 & 0.07 & 1161 & 000 go hya & + 06.0328 & 221.77 & 30.32 & 0.23 & 0.98 & 25 & 23 & 0.83 & 12.34 & 0.09 & 1141 & 111 v ind & & 355.33 & 43.12 & 7.00 & 9.00 & 202 & 3 & 1.50 & 9.92 & 0.05 & 2111 & 000 cq lac & + 39.0934 & 93.95 & 14.55 & 0.34 & 0.15 & 20 & 30 & 2.04 & 12.43 & 0.45 & 1161 & 000 rr leo & + 24.0416 & 208.42 & 53.10 & 1.69 & 1.26 & 88 & 1 & 1.57 & 10.68 & 0.09 & 1111 & 001 rx leo & + 26.0471 & 209.43 & 70.51 & 0.38 & 2.66 & 121 & 6 & 1.38 & 11.90 & 0.00 & 1111 & 000 ss leo & + 00.0760 & 265.32 & 57.06 & 2.38 & 2.79 & 163 & 3 & 1.83 & 11.03 & 0.04 & 1111 & 000 st leo & + 10.0701 & 253.44 & 66.15 & 0.56 & 3.37 & 153 & 4 & 1.29 & 11.46 & 0.09 & 1121 & 000 su leo & + 08.0618 & 228.92 & 43.82 & 0.60 & 0.86 & 81 & 30 & 1.41 & 13.55 & 0.02 & 1161 & 000 sw leo & 02.1164 & 255.63 & 48.98 & 0.74 & 0.68 & 46 & 11 & 1.45 & 13.08 & 0.07 & 1161 & 000 sz leo & + 08.0731 & 243.93 & 57.83 & 1.60 & 2.55 & 185 & 4 & 1.86 & 12.35 & 0.04 & 1121 & 000 tv leo & 05.1129 & 262.99 & 49.06 & 1.07 & 0.28 & 96 & 5 & 1.97 & 12.10 & 0.07 & 1111 & 000 ww leo & + 07.0715 & 226.04 & 38.45 & 0.03 & 2.63 & 66 & 20 & 1.48 & 12.47 & 0.09 & 1121 & 000 aa leo & + 10.0702 & 254.14 & 66.09 & 0.26 & 3.35 & 32 & 24 & 1.47 & 12.27 & 0.09 & 1121 & 000 ae leo & + 17.0930 & 234.20 & 68.19 & 2.38 & 1.25 & 53 & 10 & 1.71 & 12.52 & 0.00 & 1151 & 000 an leo & + 06.0502 & 253.35 & 60.72 & 0.29 & 3.06 & 68 & 17 & 1.14 & 12.45 & 0.13 & 1141 & 000 ax leo & + 12.0892 & 248.29 & 66.30 & 1.85 & 2.08 & 182 & 10 & 2.28 & 12.18 & 0.07 & 1141 & 000 bt leo & + 18.0551 & 228.38 & 65.59 & 0.90 & 0.34 & 119 & 14 & 0.81 & 13.11 & 0.00 & 1141 & 111 v lmi & + 29.0467 & 201.30 & 57.84 & 2.16 & 3.02 & 110 & 7 & 1.15 & 11.71 & 0.02 & 1121 & 000 x lmi & + 39.0396 & 182.53 & 53.70 & 1.63 & 2.01 & 82 & 18 & 1.68 & 12.31 & 0.00 & 1121 & 000 u lep & 21.0669 & 221.10 & 34.37 & 4.33 & 5.86 & 128 & 11 & 1.93 & 10.60 & 0.02 & 1111 & 000 ry lib & 21.1628 & 330.87 & 36.08 & 1.68 & 0.45 & 33 & 11 & 1.48 & 13.15 & 0.25 & 1161 & 000 tv lib & 08.1548 & 353.16 & 39.67 & 0.05 & 1.04 & 61 & 10 & 0.27 & 11.94 & 0.25 & 1121 & 111 vy lib & 15.2271 & 353.86 & 28.84 & 0.22 & 6.46 & 142 & 10 & 1.32 & 11.72 & 0.45 & 1111 & 000 tw lyn & + 43.0251 & 176.15 & 27.54 & 0.69 & 0.86 & 40 & 26 & 1.23 & 11.90 & 0.14 & 1141 & 101 y lyr & + 43.0988 & 72.67 & 20.87 & 0.08 & 0.66 & 65 & 23 & 1.03 & 13.28 & 0.18 & 1121 & 101 rr lyr & & 74.96 & 12.30 & 10.95 & 19.42 & 63 & 8 & 1.37 & 7.74 & 0.13 & 2111 & 000 rz lyr & + 32.1588 & 62.11 & 15.82 & 1.02 & 1.99 & 233 & 23 & 2.13 & 11.51 & 0.32 & 1111 & 000 cn lyr & + 28.1070 & 58.01 & 14.70 & 0.12 & 1.61 & 67 & 30 & 0.26 & 11.49 & 0.62 & 1161 & 111 cx lyr & + 28.1083 & 58.99 & 12.72 & 0.49 & 1.30 & 203 & 30 & 1.79 & 12.83 & 0.82 & 1141 & 000 io lyr & + 32.1543 & 60.59 & 19.98 & 1.22 & 2.19 & 157 & 30 & 1.52 & 11.86 & 0.15 & 1161 & 000 uv oct & & 308.40 & 23.55 & 6.93 & 12.30 & 126 & 12 & 1.61 & 9.42 & 0.21 & 2111 & 000 st oph & & 22.83 & 16.64 & 0.09 & 0.08 & 12 & 7 & 1.30 & 12.05 & 0.60 & 2111 & 101 v413 oph & 10.1983 & 4.39 & 25.97 & 1.10 & 1.62 & 39 & 30 & 1.00 & 12.08 & 0.63 & 1161 & 111 v445 oph & & 7.91 & 28.45 & 0.47 & 1.24 & 22 & 5 & 0.23 & 10.99 & 0.59 & 2121 & 111 v452 oph & + 11.1284 & 32.52 & 25.72 & 0.36 & 0.10 & 375 & 30 & 1.72 & 12.18 & 0.41 & 1121 & 000 v964 ori & 02.0731 & 202.50 & 23.91 & 0.70 & 1.21 & 178 & 11 & 1.89 & 12.95 & 0.25 & 1151 & 000 ty pav & & 330.55 & 17.10 & 2.10 & 2.20 & 245 & 9 & 2.31 & 12.58 & 0.26 & 2111 & 000 dn pav & & 332.82 & 30.80 & 0.90 & 3.00 & 69 & 12 & 1.54 & 12.42 & 0.14 & 2111 & 000 vv peg & + 18.1195 & 78.42 & 30.42 & 0.06 & 1.22 & 13 & 8 & 1.88 & 11.79 & 0.13 & 1111 & 000 av peg & + 22.1796 & 77.44 & 24.05 & 1.35 & 1.34 & 58 & 1 & 0.14 & 10.44 & 0.14 & 1121 & 111 bh peg & + 15.1616 & 85.62 & 38.36 & 2.01 & 6.71 & 278 & 2 & 1.38 & 10.44 & 0.20 & 1111 & 000 cg peg & + 24.0966 & 77.18 & 20.75 & 0.11 & 0.49 & 4 & 4 & 0.48 & 11.11 & 0.20 & 1121 & 111 dz peg & + 15.1715 & 93.09 & 41.46 & 1.70 & 2.49 & 294 & 11 & 1.52 & 12.00 & 0.05 & 1161 & 000 gv peg & + 26.1124 & 109.07 & 34.83 & 0.69 & 3.07 & 335 & 30 & 1.99 & 13.36 & 0.10 & 1161 & 000 ar per & & 154.93 & 2.27 & 0.15 & 0.50 & 5 & 1 & 0.43 & 10.43 & 1.08 & 2122 & 111 u pic & & 257.67 & 39.61 & 0.10 & 1.70 & 30 & 12 & 0.73 & 11.32 & 0.00 & 2131 & 111 ry psc & 02.0022 & 100.68 & 62.89 & 3.99 & 0.77 & 1 & 9 & 1.39 & 12.28 & 0.08 & 1121 & 000 bb pup & 19.0790 & 241.28 & 10.27 & 1.58 & 1.05 & 98 & 9 & 0.57 & 12.17 & 0.45 & 1111 & 111 v440 sgr & & 15.31 & 19.20 & 0.20 & 5.00 & 62 & 1 & 1.47 & 10.24 & 0.36 & 2121 & 000 ru scl & & 41.53 & 78.86 & 5.63 & 2.04 & 38 & 8 & 1.25 & 10.21 & 0.03 & 2111 & 000 vy ser & + 01.1004 & 6.16 & 44.09 & 9.84 & 0.51 & 145 & 1 & 1.82 & 10.13 & 0.06 & 1111 & 000 an ser & + 13.1114 & 23.80 & 45.24 & 0.27 & 1.11 & 47 & 4 & 0.04 & 10.97 & 0.09 & 1121 & 111 ar ser & + 02.1454 & 7.89 & 44.25 & 3.66 & 0.60 & 132 & 4 & 1.78 & 11.85 & 0.07 & 1131 & 000 at ser & + 08.1225 & 18.03 & 42.45 & 0.14 & 1.48 & 58 & 11 & 2.05 & 11.45 & 0.08 & 1111 & 000 av ser & + 00.1096 & 11.28 & 36.83 & 0.80 & 0.65 & 45 & 13 & 1.20 & 11.40 & 0.26 & 1111 & 101 aw ser & + 15.1229 & 28.67 & 43.35 & 0.97 & 1.63 & 126 & 15 & 1.67 & 12.79 & 0.04 & 1141 & 000 bh ser & + 19.0930 & 27.56 & 56.27 & 0.78 & 1.51 & 113 & 11 & 1.59 & 12.85 & 0.08 & 1161 & 000 cs ser & + 03.1110 & 7.22 & 45.43 & 2.37 & 2.79 & 2 & 14 & 1.57 & 12.39 & 0.08 & 1141 & 000 df ser & + 18.0911 & 26.27 & 55.92 & 0.46 & 0.55 & 10 & 14 & 0.74 & 12.69 & 0.07 & 1141 & 111 rv sex & 08.0980 & 258.12 & 43.38 & 0.98 & 0.85 & 120 & 20 & 1.10 & 12.30 & 0.02 & 1161 & 000 ss tau & + 05.0288 & 180.09 & 38.53 & 0.77 & 0.28 & 11 & 10 & 0.28 & 12.50 & 0.49 & 1121 & 111 u tri & + 33.0093 & 137.89 & 27.24 & 0.89 & 1.30 & 6 & 23 & 0.79 & 12.60 & 0.10 & 1121 & 111 w tuc & & 301.66 & 53.72 & 0.30 & 0.20 & 63 & 3 & 1.64 & 11.43 & 0.00 & 2111 & 000 yy tuc & & 325.32 & 54.21 & 0.14 & 0.34 & 56 & 9 & 1.82 & 11.98 & 0.00 & 2111 & 000 rv uma & + 54.0419 & 109.75 & 62.06 & 2.76 & 4.69 & 183 & 9 & 1.19 & 10.78 & 0.01 & 1111 & 000 tu uma & + 30.0521 & 198.80 & 71.87 & 7.64 & 4.97 & 88 & 1 & 1.44 & 9.81 & 0.00 & 1111 & 000 ab uma & + 48.0617 & 141.04 & 67.86 & 1.30 & 2.00 & 56 & 26 & 0.72 & 10.80 & 0.00 & 1141 & 111 st vir & 00.1211 & 346.37 & 53.65 & 0.06 & 2.65 & 22 & 13 & 0.88 & 11.52 & 0.07 & 1121 & 111 uu vir & & 280.73 & 60.52 & 2.96 & 0.45 & 8 & 1 & 0.82 & 10.56 & 0.01 & 2121 & 111 uv vir & + 00.0808 & 286.55 & 62.28 & 2.63 & 1.80 & 99 & 11 & 1.19 & 11.83 & 0.02 & 1121 & 000 wy vir & 06.1416 & 321.78 & 54.28 & 1.74 & 1.17 & 181 & 10 & 2.84 & 13.38 & 0.03 & 1161 & 000 ad vir & 07.2115 & 333.18 & 51.21 & 1.97 & 0.60 & 134 & 14 & 1.15 & 13.04 & 0.04 & 1151 & 000 ae vir & + 04.0956 & 351.61 & 57.27 & 0.03 & 1.77 & 208 & 10 & 1.16 & 13.26 & 0.02 & 1141 & 000 af vir & + 06.0757 & 355.48 & 59.16 &
6.16 & 0.05 & 35 & 14 & 1.46 & 11.52 & 0.01 & 1121 & 000 am vir & 16.1465 & 313.94 & 45.52 & 0.16 & 5.15 & 99 & 24 & 1.45 & 11.49 & 0.14 & 1121 & 000 as vir & 09.1409 & 303.47 & 52.61 & 1.14 & 3.69 & 70 & 23 & 1.49 & 11.90 & 0.08 & 1121 & 000 at vir & 05.1349 & 304.66 & 57.40 & 5.42 & 1.76 & 346 & 8 & 1.91 & 11.27 & 0.04 & 1121 & 000 av vir & + 09.0882 & 325.01 & 70.82 & 0.64 & 3.38 & 152 & 4 & 1.32 & 11.78 & 0.00 & 1121 & 000 bq vir & 02.1373 & 295.38 & 60.23 & 0.27 & 1.39 & 129 & 9 & 1.32 & 12.48 & 0.03 & 1161 & 000 do vir & 05.1546 & 345.60 & 48.45 & 2.72 & 0.69 & 24 & 36 & 0.80 & 14.14 & 0.10 & 1151 & 000 fu vir & + 13.0858 & 290.13 & 75.56 & 1.33 & 0.77 & 90 & 8 & 1.17 & 12.63 & 0.07 & 1161 & 000 fk vul & + 22.1711 & 67.60 & 13.92 & 0.05 & 1.51 & 76 & 30 & 0.95 & 12.87 & 0.35 & 1161 & 111 at and & & 109.76 & 18.09 & 0.20 & 4.60 & 241 & 11 & 0.97 & 10.66 & 0.38 & 2221 & 000 s ara & & 343.38 & 12.45 & 2.34 & 1.53 & 172 & 13 & 1.43 & 10.67 & 0.36 & 2231 & 000 ru boo & + 23.0728 & 30.94 & 63.87 & 1.35 & 0.32 & 60 & 35 & 1.50 & 13.60 & 0.04 & 1321 & 000 bi cen & & 294.66 & 2.44 & 0.76 & 0.15 & 210 & 30 & 0.83 & 11.86 & 0.59 & 2262 & 000 uu cet & 17.0006 & 73.26 & 75.09 & 2.68 & 0.65 & 114 & 3 & 1.32 & 11.95 & 0.01 & 1221 & 000 z com & + 18.0747 & 328.12 & 80.58 & 0.77 & 1.85 & 50 & 35 & 1.50 & 13.73 & 0.03 & 1321 & 000 st com & & 347.87 & 81.25 & 3.61 & 3.57 & 68 & 7 & 1.26 & 11.38 & 0.04 & 2221 & 000 sw cru & & 296.49 & 1.91 & 1.07 & 0.19 & 23 & 30 & 0.54 & 12.33 & 1.18 & 2262 & 111 uy cyg &
& 74.54 & 9.63 & 0.13 & 0.76 & 2 & 6 & 1.03 & 11.05 & 0.22 & 2222 & 101 sw her & + 21.1016 & 41.68 & 34.00 & 1.11 & 0.04 & 130 & 35 & 1.50 & 14.14 & 0.21 & 1321 & 000 vx her & + 18.0988 & 35.22 & 39.08 & 4.70 & 1.66 & 377 & 3 & 1.52 & 10.62 & 0.18 & 1211 & 000 ar her & + 47.1123 & 74.10 & 48.20 & 6.53 & 1.24 & 349 & 8 & 1.40 & 11.18 & 0.04 & 1211 & 000 rv leo & & 232.37 & 51.14 & 0.50 & 1.30 & 0 & 35 & 1.50 & 13.85 & 0.08 & 2321 & 000 tt lyn & + 44.0496 & 176.07 & 41.65 & 8.41 & 4.01 & 67 & 1 & 1.76 & 9.87 & 0.03 & 1261 & 000 ez lyr & & 65.52 & 16.25 & 1.32 & 0.20 & 60 & 23 & 1.56 & 11.60 & 0.21 & 2261 & 001 ao peg & + 18.1149 & 69.90 & 22.60 & 0.31 & 2.94 & 115 & 35 & 0.92 & 12.83 & 0.19 & 1261 & 000 tu per & & 142.78 & 4.29 & 1.51 & 0.61 & 377 & 11 & 1.50 & 12.53 & 1.38 & 2324 & 000 rv phe & & 336.01 & 64.00 & 4.15 & 1.85 & 99 & 2 & 1.60 & 11.75 & 0.06 & 2211 & 000 xx pup & & 236.65 & 8.72 & 3.13 & 0.14 & 386 & 7 & 1.50 & 11.20 & 0.38 & 2324 & 000 v675 sgr & & 358.26 & 7.83 & 0.00 & 1.20 & 105 & 30 & 2.01 & 10.36 & 0.19 & 2212 & 000 v1640 sgr & & 0.47 & 13.64 & 0.40 & 0.90 & 41 & 10 & 0.54 & 12.68 & 0.31 & 2251 & 111 v494 sco & & 357.23 & 0.49 & 0.34 & 0.62 & 26 & 30 & 1.01 & 11.27 & 1.16 & 2233 & 101 af vel & & 284.16 & 8.60 & 5.90 & 1.75 & 236 & 16 & 1.64 & 11.34 & 0.61 & 2214 & 000 bb vir & + 06.0723 & 340.31 & 64.84 & 3.71 & 1.02 & 38 & 13 & 1.61 & 11.07 & 0.00 & 1221 & 000 bc vir & + 06.0660 & 323.42 & 67.52 & 1.55 & 2.81 & 4 & 13 & 1.50 & 12.21 & 0.00 & 1361 & 000 bn vul & & 58.63 & 3.41 & 4.85 & 3.80 & 267 & 4 & 1.52 & 11.08 & 1.36 & 2262 & 000 cl disk-1 & all stars lying above / rightward of @xmath273[fe / h]@xmath274 ( see fig .
halo-1 & all stars lying below / leftward of this line .
& disk-2 & all stars having [ fe / h ] @xmath275 and @xmath276 km s@xmath5 .
halo-2 & all stars excluded from disk-2 .
& disk-3 & all stars in disk-1 @xmath277 all stars having @xmath278 km s@xmath5 , @xmath276 km s@xmath5 , & @xmath279 km s@xmath5 , @xmath280 @xmath281 kpc , @xmath221 [ fe / h ] @xmath282 .
halo-3 & all stars excluded from disk-3 .
& halo-1r & all stars in halo-1 with [ fe / h ] @xmath283 .
halo-1p & all stars in halo-1 with [ fe / h ] @xmath284 .
h1 & halo & random & 165 & 5 & h1.0 & 0.0 & yes & 5 & & & & & h1 & 0.1 & yes & 5 h2 & halo & real & 169 & 5 & h2.0 & 0.0 & yes & 5 & & & & & h2 & 0.1 & yes & 5 & & & & ( + 15 ) & h2d & 0.1 & no & 20 & & & & & & & & d1 & disk & random & 50 & 20 & d1 & 0.1 & yes & 19 d2 & disk & real & 45 & 20 & d2 & 0.1 & yes & 18 & & & & & d2d & 0.1 & no & 18 crrrrrrrrrrrr h1.0 & mean&203 & 158 & 100 & 89 & + 0.82 & mean & + 3 & 2 & 1 & 3 & + 0.05 & @xmath138 & 11 & 11 & 8 & 7 & 0.12 & sd & 5 & 2 & 7 & 3 & 0.07 & & & & & & & & & & & & h1 & mean&204 & 158 & 99 & 89 & + 0.79 & mean & + 2 & 2 & 3 & 3 & + 0.02 & @xmath138 & 12 & 11 & 8 & 7 & 0.12 & sd & 5 & 2 & 6 & 3 & 0.07 & & & & & & & & & & & & h2.0 & mean&198 & 155 & 96 & 86 & + 0.84 & mean & + 7 & 7 & 3 & 2 & + 0.07 & @xmath138 & 11 & 12 & 7 & 6 & 0.12 & sd & 2 & 5 & 3 & 4 & 0.08 & & & & & & & & & & & & h2 & mean&199 & 155 & 94 & 86 & + 0.82 & mean & + 6 & 6 & 4 & 2 & + 0.05 & @xmath138 & 12 & 12 & 7 & 6 & 0.12 & sd & 2 & 5 & 3 & 3 & 0.08 & & & & & & & & & & & & h2d & mean&205 & 155 & 93 & 87 & + 0.81 & mean & + 4 & 4 & 4 & 2 & + 0.04 & @xmath138 & 12 & 12 & 7 & 6 & 0.12 & sd & 5 & 6 & 3 & 3 & 0.08 & & & & & & & & & & & & & & & & & & & & & & & & d1 & mean & 32 & 51 & 48 & 29 & + 0.88 & mean & 0 & 2 & 1 & 2 & 0.02 & @xmath138 & 8 & 8 & 8 & 6 & 0.34 & sd & 5 & 7 & 7 & 7 & 0.44 & & & & & & & & & & & & d2 & mean & 33 & 51 & 48 & 27 & + 1.04 & mean & + 1 & 4 & 4 & 2 & + 0.13 & @xmath138 & 9 & 9 & 8 & 6 & 0.35 & sd & 5 & 6 & 6 & 5 & 0.29 & & & & & & & & & & & & d2d & mean & 32 & 51 & 48 & 26 & + 1.04 & mean & + 1 & 4 & 4 & 3 & + 0.13 & @xmath138 & 9 & 8 & 8 & 6 & 0.34 & sd & 5 & 5 & 7 & 4 & 0.29 lrcrrrrrrrr halo-1 & 169 & 1.60 & + 9 & 205 & 11 & 165 & 102 & 95 & + 0.71 & + 0.71 & & & @xmath75 13 & 12 & 8 & 12 & 7 & 7 & 0.12 & 0.12 & & & & & & & & & & halo-2 & 175 & 1.58 & + 8 & 196 & 11 & 161 & 108 & 93 & + 0.72 & + 0.72 & & & @xmath75 13 & 12 & 7 & 12 & 8 & 7 & 0.12 & 0.12 & & & & & & & & & & halo-3 & 162 & 1.61 & + 9 & 210 & 12 & 168 & 102 & 97 & + 0.71 & + 0.71 & & & @xmath75 14 & 12 & 8 & 13 & 8 & 7 & 0.12 & 0.12 & & & & & & & & & & & & & & & & & & & & halo-1r & 86 & 1.34 & 13 & 216 & 13 & 172 & 92 & 89 & + 0.69 & + 0.69 & & & @xmath75 19 & 16 & 10 & 17 & 9 & 9 & 0.16 & 0.16 & & & & & & & & & & halo-1p & 83 & 1.86 & + 31 & 195 & 10 & 154 & 111 & 100 & + 0.73 & + 0.73 & & & @xmath75 18 & 17 & 12 & 16 & 12 & 10 & 0.18 & 0.18 & & & & & & & & & & & & & & & & & & & & disk-1 & 44 & 0.66 & + 4 & 34 & 18 & 45 & 43 & 25 & + 1.24 & + 1.08 & & & @xmath75 8 & 8 & 6 & 8 & 8 & 6 & 0.34 & 0.34 & & & & & & & & & & disk-2 & 38 & 0.58 & + 8 & 43 & 19 & 51 & 47 & 25 & + 1.15 & + 1.01 & & & @xmath75 9 & 10 & 6 & 9 & 9 & 6 & 0.35 & 0.35 & & & & & & & & & & disk-3 & 51 & 0.76 & + 6 & 45 & 16 & 52 & 48 & 29 & + 0.94 & + 0.79 & & & @xmath75 8 & 9 & 6 & 8 & 8 & 5 & 0.30 & 0.30 & & & & & & & & & & all stars & 213 & 1.40 & + 7 & 169 & 14 & 147 & 115 & 85 & + 0.73 & + 0.73 & & & @xmath75 10 & 10 & 6 & 10 & 7 & 6 & 0.11 & 0.11 | we present new statistical parallax solutions for the absolute magnitude and kinematics of rr lyrae stars .
we have combined new proper motions from the lick northern proper motion program with new radial velocity and abundance measures to produce a data set that is 50% larger , and of higher quality , than the data sets employed by previous analyses . based on an _ a priori _ kinematic study , we separated the stars into halo and thick disk sub - populations .
we performed statistical parallax solutions on these sub - samples , and found @xmath0 at @xmath1[fe / h]@xmath2 = 1.61 for the halo ( 162 stars ) , and @xmath3 at @xmath1[fe / h]@xmath2 = 0.76 for the thick disk ( 51 stars ) .
the solutions yielded a solar motion @xmath4 km s@xmath5 and velocity ellipsoid @xmath6 km s@xmath5 for the halo .
the values were @xmath7 km s@xmath5 and @xmath8 km s@xmath5 for the thick disk .
both are in good agreement with estimates of the halo and thick disk kinematics derived from both rr lyrae stars and other stellar tracers .
monte carlo simulations indicated that the solutions are accurate , and that the errors may be smaller than the estimates above .
the simulations revealed a small bias in the disk solutions , and appropriate corrections were derived .
the large uncertainty in the disk @xmath9 prevents ascertaining the slope of the @xmath9[fe / h ] relation . using a zero point defined by our halo solution and adopting a slope of 0.15 mag dex@xmath5
, we find that ( 1 ) the distance to the galactic center is @xmath10 kpc ; ( 2 ) the mean age of the 17 oldest galactic globular clusters is @xmath11 gyr ; and ( 3 ) the distance modulus of the lmc is @xmath12 mag .
estimates of @xmath13 which are based on an lmc distance modulus of 18.50 ( e.g. , cepheid studies ) increase by 10% if they are recalibrated to match our lmc distance modulus . |
modified newtonian dynamics , mond , ( milgrom 1983 ) is a non - relativistic theory that posits the existence of a critical acceleration ( @xmath0 cm / s@xmath1 ) below which the effective gravitational acceleration @xmath2 deviates from newtonian form ( @xmath3 ) in a specific way in effect , @xmath4 when @xmath5 .
the motivation is to remove the need for dark matter in gravitationally bound astronomical systems with low internal and external accelerations ( @xmath5 ) .
the critical acceleration can also be expressed as a surface density ( @xmath6 ) , the implication being that discrepancies between the classical dynamical mass and the observable baryonic mass should appear in low surface density , or low surface brightness , systems .
conversely , there should be no significant discrepancy within high surface brightness systems rotation curves of the neutral gas in disk galaxies as measured in the 21 cm line present obvious advantages in tracing the gravitational acceleration as a test of mond : cool gas , generally in planar circular motion , provides an unambiguous tracer of the acceleration and deviations from such motions can usually ( but not always ) be identified ; moreover , random motions are small and generally do not contribute to the support of the gas disk against gravity ( trachternach et al . 2008 ) .
the success of mond when confronted by the extensive body of data on measured rotation curves , ranging from gas rich low surface brightness dwarfs ( swaters et al .
2010 ) to high surface brightness earlier type disk galaxies dominated by a stellar component ( sanders and noordermeer 2007 ) , can hardly be disputed .
this success , plus the observed and theoretically predicted baryonic tully - fisher relation ( mcgaugh 2005 ) constitute the principal evidence supporting mond .
however , gas poor early type systems , ellipticals and s0s , usually miss such a clear tracer of the gravitational acceleration ; for such objects the situation has been more confused .
an unavoidable prediction of mond is that in high surface brightness systems , such as luminous elliptical galaxies , there should be little discrepancy between the detectable baryonic mass and the newtonian dynamical mass within the bright luminous object . in other words , with the traditional newtonian analysis , there should be no evidence for dark matter within the projected radius containing half the flux of visible light , the effective radius .
this , in fact , was the result of the observational study of romanowsky et al .
they used the observed kinematics of bright planetary nebulae as a tracer of the mass distribution in three nearby elliptical galaxies , and found that the results were consistent with no substantial dark matter contribution within four effective radii , a result shown by milgrom and sanders ( 2003 ) to be in agreement with the expectations of mond .
the use of such stellar tracers suffers from the ambiguity introduced by the uncertain distribution of stellar orbits , but milgrom ( 2012 ) has recently demonstrated that the pressure distribution of the hot x - ray emitting gaseous envelopes in two isolated elliptical galaxies , extending out to 100 kpc and over a range of a factor 100 in acceleration , is entirely consistent with the run of gravitational accelerations calculated from the observed distribution of visible stars using the mond algorithm .
the implied mass - to - light ratios of the stellar populations are sensible for early type galaxies . subsequently , milgrom ( 2013 ) pointed out that the statistics of galaxy - galaxy weak gravitational lensing ( the small distortions in the shapes of background galaxies in the field of nearer foreground galaxies ) , which probes accelerations down to a few percent of @xmath7 , implies the asymptotic velocity dispersion of fitted isothermal spheres is related to the baryonic mass of the deflecting galaxies as @xmath8 , exactly as required by mond
an elliptical galaxy equivalent to the baryonic tully - fisher relation in spiral galaxies .
in other words it appears that the dynamics of the outer halos " , more than one hundred kiloparsecs in extent , is determined by the small fraction of baryons in the very center , a very strange fact indeed when viewed in the context of dark matter .
on the other hand , there are persistent claims that strong gravitational lensing , the formation of multiple images or einstein rings of background sources by foreground galaxies , require the presence of substantial quantity of dark matter within one or two effective radii in early type galaxies .
if true , this would appear to be in contradiction to the predictions of mond because strong lensing can only occur in the high acceleration regime ( see the discussion below ) .
there has been controversy about this issue in the literature , with some authors claiming that no dark matter is required ( chiu et al .
2008 , chen and zhao 2008 , sanders and land 2008 , chiu et al .
2011 ) , while others others argue strong lensing requires that substantial fraction of the total mass ( up to 80% ) is dark within two effective radii ( mavromatos et al .
2009 , ferreras et al .
2009 , leier et al .
2011 , ferreras et al . 2012 ) .
this problem is complicated by the possibility of contamination whether or not distant lenses are truly isolated or lying within groups or clusters and by the uncertainty of the mass - to - light ratio of the underlying stellar population . with respect to this second problem that of the stellar m / l there is recent evidence , spectroscopic and dynamical , that the initial mass function of stars formed in early - type galaxies , the imf , is not universal , as is often supposed , but becomes increasingly bottom heavy weighted toward lower mass stars in higher mass galaxies .
that is to say , the imf is better described by that of salpeter ( 1955 ) rather than that of chabrier(2003 ) in systems with higher velocity dispersion or total mass .
this in turn implies that the total stellar mass - to - light ratio is higher in more massive galaxies ( there is about a factor of two difference in m / l between model populations constructed with the chabrier vs. salpeter imfs ) .
spectroscopic evidence for an increasingly dominant contribution of dwarf stars has been given by conroy and van dokkum ( 2012 ) , smith et al .
( 2012 ) and spiniello et al .
( 2012 , 2013 ) .
conroy et al .
( 2013 ) have provided dynamical evidence that in compact elliptical galaxies ( in which dark matter presumably does not dominate ) the stellar m / l increases systematically with galaxy velocity dispersion . the evidence of spiniello et al .
( 2013 ) , using spectroscopic tracers of low mass stars , similarly suggests an almost linear increase in the stellar m / l with velocity dispersion . here ,
in view of these developments , i reconsider the question of whether or not the mond lensing masses of strong lens systems are consistent with the stellar masses in the slacs sample of gravitational lenses .
unlike general relativity ( gr ) mond is a non - relativistic theory .
that means that mond in itself says nothing about gravitational lensing and other relativistic effects .
there have now been several proposed candidate relativistic extensions of mond ( see famaey and mcgaugh 2011 , for a recent review ) but there is no generally accepted theory . in most of these proposals ,
the relationship between the deflection of photons and the weak field force , also in the low acceleration limit , is required to be the same as it is in gr .
this is built into the theories and , in fact , required by apparent coincidence of the classical dynamical mass of clusters of galaxies ( using galaxy kinematics or the distribution of hot gas non - relativistic particles ) and the lensing mass of the clusters ( the effect of gravity on photons relativistic particles ) .
therefore , it is not yet particularly meaningful to apply a specific relativistic theory , such as teves ( bekenstein 2004 ) , to the problem of lensing ; the suggested interpolating functions are just as arbitrary as those typically used in mond .
so here i will use one of the usual mond interpolating functions that works well for galaxy rotation curves , the so - called simple form recommended by zhao & famaey ( 2006 ) ; i.e. , @xmath9 with @xmath10 .
the mond algorithm is given by @xmath11 where i take @xmath12 cm / s@xmath1 .
viewing mond as a modification of gravity , this formula is only strictly true in spherical systems .
it can be shown that , with gr , for a spherically symmetric gravitational lens , perfectly aligned with a small background source , an einstein ring will be formed if the enclosed mean surface density in the lens exceeds @xmath13 where @xmath14 is a dimensionless function of the lens and source redshifts ( related to the angular size distances in units of the hubble distance , @xmath15 ) ; for the lenses considered here @xmath16 . the critical surface density below which mond phenomenology appears is roughly @xmath17 given that @xmath18 we find that , typically @xmath19 therefore strong lensing always occurs in the high acceleration limit .
no large discrepancy should be detected by strong gravitational lensing ( the minimum value of @xmath20 in a concordance cosmology is 3.4 , so it is always the case that @xmath21 ) .
that is not to say that no discrepancy whatsoever should be present within an einstein ring radius .
mond can be represented by a halo of phantom dark matter
the dark matter that one would presume to be present if the mond phenomenology were to be represented by a dark halo .
the space density of phantom dark matter is given by @xmath22 where @xmath2 is the mond gravitational acceleration given by eq . 2 and @xmath23 is the density of detectable baryonic matter .
asymptotically , the mond acceleration ( the solution determined by eq . 2 about a point mass ) goes as @xmath24 then eq . 6 implies that in the outer regions @xmath25 as in an isothermal sphere . because this phantom halo is seen in projection
, it will contribute roughly 15% of the total _
projected _ mass within an einstein ring ( the exact fraction depends upon the interpolating function @xmath26 ) . of course , phantom dark matter is a phantom , but for determining the lensing properties of an object with mond , the concept is useful .
for example , the mond critical surface density for strong lensing is identical to that given by eq .
3 _ when the projected phantom dark matter is included _ ; the true surface density of projected baryonic matter is reduced by this same factor ( @xmath27 15% ) .
the slacs lenses ( sloan lens advanced camera surveys ) comprise a reasonably large ( 85 ) , homogeneously selected sample of strong gravitational lenses at redshifts typically between 0.1 and 0.3 .
most of the objects are early type galaxies , elliptical or s0 , and a number of these present almost complete einstein rings , simplifying the modeling and leading to quite unambiguous lensing mass estimates inside the einstein ring radius .
the sdss observations also include measurements of the stellar velocity dispersions within an aperture of three arc seconds .
auger et al .
( 2009 ) have imaged the lensing galaxies with the hst in various photometric bands , and the effective radius of the corresponding de vaucouleurs profile is determined .
the broad band colors permit the fitting of stellar population synthesis models and , thereby , estimates of the stellar mass .
an important free function in these models is the form of the imf , and auger et al .
have considered both the popular chabrier and salpeter forms ; they tabulate the estimated stellar masses of the lensing galaxies in both cases . [ cols="<,^,^,^,^,^,^,^,^,^ " , ] here i have selected 65 objects from their sample .
these are lenses which are classified as elliptical galaxies , and which have complete photometric data , all with estimates of the stellar mass . in the dynamical analysis
i assume that the total light and mass distribution is given by the spherical jaffe model ( jaffe 1983 ) with an effective radius appropriate to the particular object .
the effective radius is that provided by auger et al .
based upon de vaucouleurs law fits to the i - band photometry ; i take this to be representative of the true distribution of starlight and stellar mass ( the scale length of the corresponding jaffe model is given by @xmath28 ) .
given the numerical value of @xmath14 ( calculated in the context of the standard concordance " cosmology which the proper theory of mond should reproduce ) and the effective radius in each case , i adjust the mass of the jaffe model , for both gr and mond , in order to match the einstein ring radius .
that is the radius ( also given by auger et al .
2009 ) within which the enclosed surface density , is equal the critical surface density ( eq .
3 ) which , in the case of mond , includes projected phantom dark matter ( eq .
the projected jaffe model mass within the einstein ring radius is the lensing mass for gr or mond , but with mond the lensing mass is , on average , 15% lower because of the higher effective gravitational force .
i emphasize again that the difference between the gr and mond lensing mass is equal to the contribution of projected phantom dark matter .
it is the lensing mass in both cases that i will compare with the stellar mass projected within the einstein ring assuming that the jaffe model with the i - band effective radius describes the luminosity density .
one could alternatively choose to compare the total jaffe mass either for gr or mond with the total stellar mass , but this obscures the fact that strong lensing provides only a measurement of the projected mass within the einstein ring .
the observed and derived parameters of the slacs subsample are given in table 1 . here for each lens i give 1 ) the einstein ring radius in kpc ; 2 ) the i - band effective radius in kpc ; 3 ) the total mass of the jaffe sphere required to produce the observed einstein ring radius in the context of mond ( in all cases masses are given in units of @xmath29 ) ; 4 ) the projected mass within the einstein ring , the lensing mass , with gr ( this is consistent with that given by auger et al .
2009 , table 4 column 3 ) ; 5 ) the projected jaffe model mass within the einstein ring in the context of mond , the mond lensing mass ( the difference between columns 4 and 5 is the contribution of phantom dark matter ) ; 6 ) the stellar mass ( salpeter ) projected within the einstein ring ( this depends upon the effective radius ) ; 7 ) the fraction of the mond mass to gr lensing mass ; 8) the fraction of projected visible to gr lensing mass ( ratio of column 6 to column 4 ) within the einstein ring ; 9 ) the total mond mass - to - light ratio in the visible band .
note that column 3 provides the normalization of the jaffe model in the context of mond .
the jaffe mass normalization for gr is obtained by multiplying the mond jaffe mass ( column 3 ) by the ratio of the gr lensing mass to the mond lensing mass ( column 4 to column 5 ) . from the table
we see that with mond there is rather little dispersion of the jaffe mass fraction within the einstein radius ( column 7 ) : @xmath30 .
this is because the einstein radius depends upon the interior surface density as does the radius of the onset of modified dynamics . with mond
this would be the true baryonic mass in terms of the estimated gr lensing mass .
on the other hand the fraction of stellar mass to gr lensing mass , @xmath31 , shows a greater dispersion , with @xmath32 . the dispersion in this quantity , also given by auger et al .
2009 ( table 4 , column 5 ) , reflects the errors , random and systematic , in the population synthesis models for the stellar mass .
in general the results for @xmath31 given here correlate with those of auger et al .
( 2009 ) although their mean value is lower ( @xmath33 ) implying a larger fraction of dark matter .
this is because of the use here of the i - band @xmath34 as opposed to the v - band by auger et al .
the effective radii in v are generally larger which means a smaller stellar mass projected within the fixed einstein ring radius .
i comment on this in the final section .
the lensing mass with gr ( column 4 ) is plotted against the projected stellar mass ( salpeter ) in fig . 1 , and the mond lensing mass against the projected stellar mass in fig .
2 . as noted above , the mond lensing mass is lower because of the enhanced deflection due to the larger effective gravitational force .
the quantities being plotted here are completely independent : the stellar mass is determined from stellar population models based upon the observed colors of the lenses , and the lensing masses are determined by the einstein ring radius with the assumed law of gravity
. a quantitative measure of the discrepancy in these objects would be the ratio of the projected to stellar mass . for newton ,
the mean value of this ratio over the sample is 1.21 @xmath35 0.27 ; with mond this is 1.03 @xmath35 0.22 .
it is evident that both determinations of the dynamical mass are consistent with each other and with the presence of no discrepancy , as mond would predict .
when the stellar masses are estimated using the chabrier imf , there is an apparent discrepancy with the mean ratio of mond lensing to stellar mass being 1.90 @xmath35 0.44 .
this demonstrates the importance of the assumed imf in assertions about the contribution of dark matter within the inner regions of elliptical galaxies : a factor of two difference between estimated lensing and stellar masses can not be taken as evidence for dark matter .
m@xmath36 and the equality line is shown.,height=264 ] several of the objects included here ( table 1 ) have been considered in detail in separate studies .
for example sdss j1430 + 4105 was discussed by eichner , seitz & bauer ( 2012 ) who argued that several subcomponents in the lensed image constrain on the total mass distribution within the einstein radius and that the fractional dark mass within the einstein ring range ranges from 0.2 to 0.4 ( @xmath37 ) .
the lens is complicated by the fact that it is not isolated ; there is a surrounding group .
none - the - less , the results are roughly consistent with those given here with @xmath38 .
the total @xmath39 of this object ( with modified dynamics ) is the second largest of the sample ( 6.48 ) , but it is certainly not an outlier from the distribution of points on fig . 2 .
we see from figs . 1 and 2 that there is some evidence for a larger discrepancy at larger galaxy stellar masses ( the points appear to form a steeper relation than the equality line ) .
indeed there are claims that the imf systematically varies with galaxy mass or velocity dispersion toward being more bottom heavy ; i.e. there is a larger stellar m / l in more massive systems ( spiniello et al .
there is support for this claim in fig . 3 : here the mond lensing
mass - to - light ratios are plotted against the observed stellar velocity dispersions .
the horizontal line is the mean m / l@xmath40 in the rest frame of 4.2 @xmath35 1.0 .
the points with error bars are the m / l averaged in bins of 15 objects and the line is the relation suggested by spiniello et al . on the basis of spectroscopic tracers of low mass stars .
we see that the distribution of mond lensing mass - to - light ratios is reasonable for elliptical galaxy stellar populations and that there is marginal evidence for an increase in m / l with velocity dispersion .
1.0 ) , and the large points with error bars are the mean m / ls averaged in bins of 15 objects . the dashed line is the fit to stellar m / l as a function of velocity dispersion given by spiniello et al .
statements about the need for non - baryonic dark matter within the bright visible inner regions of strong gravitational lenses early type galaxies are not supported by the evidence given here .
figs . 1 and 2
demonstrate that the gr and mond lensing masses are consistent with each other and with the masses of the stellar components determined from population synthesis modeling using the salpeter imf .
this is in agreement with the expectation from mond : there should be little discrepancy between the visible and the newtonian lensing mass within high surface brightness early type systems ; the discrepancies only appear in the outer regions .
it is also evident that the uncertainties introduced by the assumed imf are at least a factor two . within this factor ,
no assertion about the need for dark matter based upon use of a particular stellar imf can be credible .
these conclusions are quite independent of the way in which the systems are modeled .
use of hernquist ( 1990 ) rather than jaffe models give similar results but with slightly higher ratios of lensing mass to stellar mass ( about 6% on average ) .
the hernquist model does have the advantage that in a given object the radial distribution of stellar velocity dispersion is more nearly constant ( isothermal ) as observed . since the two classes of models bracket the empirical de vaucouleurs law ( sand , et al .
2004 ) , for which the effective radius is measured , the use of an exact @xmath41 model would certainly lie within this range of 6% .
it is interesting that auger et al .
( 2010 ) claim that these observations do imply the existence of dark matter within the inner parts of ellipticals .
there are two reasons for the difference with the conclusions of the present work .
the first is that for auger et al .
the bench mark for defining the inner regions is one - half the effective radius , whereas here i take the einstein ring radius . with respect to mond
this is more appropriate because the onset of modified dynamics ( or the appearance of dark matter " ) is tied to the enclosed surface density as is the location of the einstein ring .
secondly , auger et al .
use the v - band effective radius , whereas here the i - band is taken as the indicator of the distribution of light and stellar mass . which ever is more appropriate , the differences in the estimated stellar mass within the einstein ring are not large and within the uncertainties of population synthesis modeling . most of the claims of need for dark matter within the einstein radius are based upon observations of more distant systems , such as those of the castles sample ( ferreras et al .
these lenses have a wide distribution of redshifts , with typically @xmath42 but ranging up to @xmath43 .
these are generally more complicated lens systems with multiple images and relatively few complete einstein rings ; therefore the lens modeling is less certain . moreover , for these distant lenses it is more difficult to detect contamination by background objects groups or clusters surrounding the lens or along the line of sight and it does appear that the contamination rate is higher for this sample ( leier et al .
it would seem to require substantial structural evolution since @xmath44 if the lenses in this sample really do have more dark matter in the inner regions than the objects in the relatively nearby homogenous and clean slacs sample .
overall , the mond prediction of no significant discrepancy between the newtonian dynamical mass and stellar mass within the inner high surface brightness regions is supported by this analysis of the slacs sample .
indeed , one can turn the argument around : figs . 1 and 2 support the validity of stellar population synthesis models in determining the mass of the stellar component .
i thank leon koopmans , chiara spiniello , moti milgrom , stacy mcgaugh , and marcel pawlowsky for helpful comments .
an anonymous referee suggested numerous improvements in the presentation of these results . | i show that the lensing masses of the slacs sample of strong gravitational lenses are consistent with the stellar masses determined from population synthesis models using the salpeter imf .
this is true in the context of both general relativity and modified newtonian dynamics , and is in agreement with the expectation of mond that there should be little classical discrepancy within the high surface brightness regions probed by strong gravitational lensing .
there is also dynamical evidence from this sample supporting the claim that the mass - to - light ratio of the stellar component increases with the velocity dispersion . |
there is a class of processes in which the competition plays a very important role .
examples are ecological , political , epidemiological , economic , chemical , reaction - diffusion , biological systems .
an important sub - class of those is the cyclic competition systems . in ecology ,
cases when variants of a species compete with one - another in a cyclic fashion have been observed @xcite .
another system of interest are cyclic food webs . in politics ,
different political parties compete and replace one - another in the helm of power . in the epidemiological context , examples are diseases which do not leave permanent immunity , known otherwise as @xmath9 ( susceptible - immune - recovered - susceptible ) models @xcite .
goodwin @xcite introduced a system of interacting biochemical metabolic oscillators , which has an autocatalytic feedback mechanism .
biochemical reactions in a cell support its activities , hence assuring its very existence .
autocatalytic reactions are an important class of reactions within a cell .
they are reputed to have made possible the birth and existence of life itself .
the simplest example of an autocatalytic reaction is the loop of the type @xmath10 , where @xmath11 .
the molecules are in a well - stirred container ( the cell ) , which is in contact with a reservoir ( the outside environment )
. they can migrate into and out of the container , to and from the reservoir . in another ( ecological ) context , the @xmath12 s are versions of a biological species , and in the epidemiological one , states of an individual ( e.g. susceptible , infected , etc . ) in ecological systems it makes sense to also study the neutral version of this model , in which @xmath13 or @xmath14 with equal probability , corresponding to the kimura model of neutral genetic drift @xcite .
duty @xcite has worked on the two allele almost neutral drift model with mutations .
the almost neutral model with mutations , preserving the total number of individuals , has only one degree of freedom , and allows one to derive an `` effective potential '' from the fokker - planck equation , obtained by a kramers - moyal expansion of the master equation . for small mutation probabilities , such that @xmath15 , there is extinction of one species and fixation .
the effective potential is almost symmetric around the centre ( where both species are in equal numbers ) and the branches of the effective potential are down .
this allows for the system to quickly slip into a state where only one of the species is present . otherwise , both species coexist forever in the high mutation regime , i.e. when @xmath16 . in that regime ,
the effective potential is symmetric around the centre point , but with branches upwards , which means that the centre is a minimum potential point .
the system then remains in the vicinity of that point for very long times .
the effective potential `` flips '' from `` branches up '' to `` branches down '' at the point where @xmath17 .
the transition is second - order , and critical behaviour is observed . in two previous studies @xcite
we have considered an @xmath18 model with cyclic competition / neutral drift and mutations ( migrations ) at a constant probability .
the system exhibits a critical transition from a `` fixation '' regime to a `` neutral '' one , in which biodiversity is preserved over long time . in the `` fixation '' regime ,
the number of the @xmath19 species oscillates with an amplitude that drifts with time , until one of the species ( and then the second one ) goes extinct , except for occasional `` bursts '' ( which are absent when there is no migrations in the picture ) . in the `` diversity '' ( or `` neutral '' ) regime , the number of the @xmath19 versions fluctuates around the centre point , and there are rare extinctions , but the product @xmath18 remains nonzero almost always .
the survival probability decays exponentially below the transition point , but the exponent decreases as the mutation ( migration ) probability per particle increases , until it becomes zero at the critical point . the critical mutation ( migration ) probability depends on system size as @xmath20 , and the models have the same power - law exponent : @xmath21 .
there is no qualitative difference between the system with mutations and that with migrations . in the present paper
we study the system with four or more species , and show that the above - described picture holds .
we show that the cyclic system is a generalisation of the well - known lotka - volterra system , and that the size - induced transition is present in those systems as well .
our system is an autocatalytic loop of the type @xmath22 , where @xmath23 .
the molecules are in a well - stirred container ( the cell ) , which is in contact with a reservoir ( the outside environment ) .
they can migrate into and out of the container , according to the following rules : a molecule ( individual ) of species @xmath24 leaves the container at a rate @xmath25 , and enters it at a rate @xmath26 , where @xmath27 and @xmath28 are its concentrations in the cell and the reservoir , respectively .
( we will assume for simplicity that the reservoir is large enough , so that the exchanges do not perturb it . )
the rate equations then read : @xmath29 in the rate equations approximation , the system size is conserved . the above equations ( [ eq : rateqm ] )
have a fixed point , and it is a stable solution .
the above model relates to the famous lotka - volterra @xcite model of interacting populations .
a very good , and pleasant to read , review of the model has been written by goel , maitra , and montroll @xcite .
volterra described the system of two kinds of fish ( predator - prey ) by the pair of equations : @xmath30 the terms of the form @xmath31 describe respectively the depletion of the stock of fish 1 and the enrichment of that of fish 2 from eating fish 1 , and those of the form @xmath32 the evolution of the species @xmath24 in absence of the other .
further on , he generalized the pair of equations ( [ eq : vol2 ] ) to @xmath8 species : @xmath33 in absence of other species , the @xmath24-th species will grow or die exponentially , depending on the sign of @xmath34 . the constants
@xmath35 will be positive , if species @xmath24 eats @xmath36 , negative , if it is eaten by @xmath36 , and zero , if they do not interact at all .
this leads to @xmath37 .
the mean - field ( rate equations ) behaviour of the system is well - known .
it accepts periodic solutions , as shown by volterra .
statistical mechanics of the volterra system was first constructed by kerner @xcite .
goel et al .
@xcite studied the existence and stability of the solutions to the rate equations .
they also introduced a random function of time ( noise ) in the growth equation , obtaining a liouville equation .
the model is indeed one of the simplest of nonlinear competition models .
jain and krishna @xcite use directed graphs to describe autocatalytic sets .
directed graphs are a set of ` nodes ' and ` links ' , where each link is an ordered pair of nodes @xcite . a graph with @xmath38 nodes
is specified by its adjacency matrix @xmath39 , defined in such a way that the @xmath40 element is one , if the graph contains a link directed from node @xmath36 to node @xmath24 and zero otherwise . for any non - negative matrix ,
the perron - frobenius theorem @xcite states that there exists an eigenvalue which is real and larger than or equal to all the other eigenvalues in magnitude .
it has been shown @xcite that when a graph has a closed loop , the perron - frobenius eigenvalue of its adjacency matrix is exactly 1 .
the set of perron - frobenius eigenvectors is of interest to us , since it provides the attractors for the dynamical system whose evolution is described by the set of differential equations : @xmath41 which is equivalent to the generalized volterra equation : @xmath42 when the catalysed reaction is much faster than the one described by the first term ( the spontaneous one ) .
the system will then converge towards the fixed point that is a perron - frobenius eigenvector of the graph s adjacency matrix . in the case of our cyclic systems ,
the perron - frobenius eigenvalue is one , and the components of the corresponding eigenvector are all equal .
this means that , in the rate equations approximation , the symmetric system ( with all the rates equal to one ) will approach the centre ( all @xmath27 s are equal ) , and remain there .
a description of the newest techniques of stability analysis for such systems can be found in @xcite , and the references therein .
however , the rate equations are just a `` mean field '' approximation ; they only describe the behaviour of the average values of the individual populations . in the real world , the system is subject to stochastic noise due to birth and death processes ( intrinsic noise ) , which we take to be poisson - distributed .
the random nature of these processes need be taken into consideration , if we want to obtain the correct and complete behaviour of the system . for
that we ought to write the master equation , and then somehow solve it .
unfortunately , very few master equations are simple enough to accept analytical solutions .
we deal with this situation by expanding them into a fokker - planck equation , which then helps us draw the necessary information about the behaviour of the system .
another approach is to simulate the master equation of the system .
the master equation for , say , the four - species cyclic system is : @xmath43@xmath44 p(a_1,a_2,a_3,a_4,t)\ ] ] where we have used the `` shift '' operators notation : @xmath45 and similarly for the other concentrations .
for systems like the one above , where the rate equations have a stable solution , the @xmath46-expansion of van kampen @xcite works exceptionally well .
previously we have used it to solve a three - species cyclic and neutral system in presence of mutations and migrations @xcite . using the `` shift '' operators notation ( [ eq : shift ] )
, the master equation for the cyclic competition system with migrations reads : @xmath47@xmath48 + d \left [ ( \epsilon_1 - 1)a_1 + ( \epsilon_2 - 1 ) a_2 + \right.\]]@xmath49 \right \ } p(\ { a_i \ } , t)\ ] ] the idea of the van kampen expansion @xcite is to split the variables of the problem into a non - fluctuating part , and a fluctuating one , i.e. deal separately with the mean - field solutions and the fluctuations ( which are taken to be of the order @xmath50 ) . in this approach , the numbers of the individual populations would be written as : @xmath51 here the @xmath52 are the steady - state ( non - fluctuating ) concentrations of the @xmath24-th species respectively ( which only depend on time ) , and the @xmath53 are the fluctuations ( proportional to the square root of system size ) . then the probability distribution @xmath54 is transformed into @xmath55 , and the following relations are true : @xmath56 @xmath57 and @xmath58 next we substitute everything into the master equation , leave only the term @xmath59 on the left hand side , and group the right hand side terms according to powers of @xmath50 .
the first term is of order @xmath60 , and it must be equal to zero , for an expansion in terms of @xmath60 to make sense .
that term reproduces the rate equations in terms of the concentrations @xmath52 , with steady state solution @xmath61 .
the terms of order @xmath62 give a linear fokker - planck equation of the form : @xmath63\ ] ] where the a - matrix for the cyclic system is : @xmath64 and for the neutral system : @xmath65 the b - matrix is the same for both systems .
its diagonal elements are @xmath66 , and the off - diagonal ones : @xmath67 .
the fokker - planck equation obtained this way is linear , and the coefficients depend on time through @xmath52 .
we are interested in fluctuations around the steady state .
this approximation is otherwise known as `` linear noise approximation '' .
the solution is known to be a gaussian ; the problem represents itself as an ornstein - uhlenbeck process . for our purposes , it suffices to determine the first and second moments of the fluctuations . following van kampen @xcite
, we can multiply the fokker - planck equation by @xmath53 and integrate by parts to get : @xmath68 for simplicity we can assume that all the concentrations in the reservoir are equal : @xmath69 .
the eigenvalues of the @xmath70 matrix are @xmath71 ( doubly - degenerate ) , @xmath72 for the cyclic system , and @xmath71 ( quadruply degenerate ) for the neutral system . here
@xmath73 is the steady state value of the concentrations ( we have dropped the index , since they are all the same ) .
the negativity of the eigenvalues guarantees the stability of the zero solutions to the first moments equations .
hence , the average of the fluctuations decays to zero and remains zero .
the equations for the second moments can be obtained similarly : @xmath74 they depend on time through @xmath52 s , which we again substitute by their steady state value , since we are interested in the fluctuations around that state . also , by symmetry , all the diagonal terms @xmath75 are equal , as well as off - diagonal terms ( correlations ) @xmath76 .
they depend on the migration probability @xmath77 alone .
the steady state solutions for the diagonal terms ( and also for the variances , since the mean values are zero ) , are as follows : @xmath78 where @xmath79 is the concentration of any of the species in the reservoir .
the number of individual species will fluctuate around @xmath80 where @xmath81 is the steady - state concentration .
all the species survive forever . this way ,
( sufficient ) migrations into and out of the container maintain diversity in the system .
it is important to point out that from the expression for the a - matrix above , we can tell that when migrations are absent ( @xmath82 ) , the ( real part of all ) eigenvalues of that matrix become zero , and the equation is of the diffusion type @xcite . in that case
, the rate equations suggest that the non - fluctuating part will remain constant ( at the centre )
. however , small deviations give rise to large differences , and one would expect the fluctuations to grow , rather than remain limited .
the separation into a macroscopic part and small fluctuations is no longer meaningful . in this case , after a transient period ( of the order @xmath60 ) , one would expect @xmath83 to be a smooth function of the concentrations , and expand in powers of @xmath84 , rather than @xmath60 .
this is otherwise known as kramers - moyal expansion , and we have employed it for the three - species systems @xcite , in very good agreement with the simulations results .
furthermore , the only difference between the three and four species systems expansion is the appearance of an `` off - diagonal '' of zeroes , in positions @xmath85 for which the species @xmath86 and @xmath87 do not react .
this means that the above algebra will remain exactly valid when there are more than four species in the system .
hence , our results will hold for any number of species , and any reasonable system size .
if there is only migrations into and out of the container ( i.e. no cyclic reactions ) , the system remains near the centre point , and the product of the concentrations remains considerably above zero ; in other words , all species are present in the system at ( almost ) all times .
when both the cyclic / neutral mechanism and migrations are present , one can occasionally observe temporary extinctions .
since the boundary is not absorbing , occasional migrations will return the system to the state with maximal symmetry ( diversity ) where all species coexist .
the migrations then manage to keep the system maximally disordered , since they are stronger than the fluctuations ( which try to drive the system toward the boundary , i.e. fixation , and keep it there ) .
the migration rate acts then as some sort of `` temperature '' , and decreasing the migration rate would be analogous to annealing the system .
in two papers , togashi and kaneko @xcite focus on the four - species autocatalytic system with a very small number of molecules .
their work covers many aspects of a transition , which they baptize `` discreteness - induced '' .
the idea of their work is that , when the number of molecules in the system is very small , a transition from the state where the numbers of molecules of the individual types are gaussian - distributed , as derived above , to one in which the @xmath4 and @xmath5 species form an alliance against species @xmath6 and @xmath7 is observed . in their work
they keep the diffusion rate @xmath77 constant , and vary the system size . as the total number of molecules ( system size ) decreases and goes through a certain value ( which coincides with @xmath88 , the inverse of migration rate ) , they observe a transition into the broken - symmetry state .
one of the symbiotic pairs eventually wins , but , since there are migrations into and out of the container happening all the time , the introduction of a member of the opposite symbiotic pair may bring the victory of that initially ( almost ) non - existent pair , making the pair that previously was the `` winner of the hour '' disappear , and so on .
the probability distribution of the number @xmath89 ( corresponding to our @xmath90 ) is shown in fig . 2 of their article @xcite , and the formation of the peaks that indicate the ( temporary ) victory of one such symbiotic pair over the other as the system size goes through @xmath88 is quite pronounced .
4 of their prl article @xcite togashi and kaneko show a plot of the rate of residence of the 1 - 3 ( or 2 - 4 ) rich state ( i.e. the state in which one of the pairs `` rules '' ) as a function of the product @xmath91 ( @xmath92 is system size , the parameter we call @xmath84 ) .
the rate of residence of the symmetry - broken state clearly goes to zero , as @xmath93 .
looking at the expression for the variance of the fluctuations above ( [ eq : solmig ] ) , one can observe that when the migration probabilities per particle ( migration rate ) approach zero , the variance of the concentrations of individual populations is of the order @xmath94 , and it becomes of order 1 ( i.e. the order of macroscopic concentrations , ) when @xmath95 ( here we are using the values of parameters as chosen by togashi and kaneko @xcite , who assume all @xmath96 .
however , it works exactly the same way with our steady state values @xmath97 ) .
this gives us the critical value for the migration probability .
the critical @xmath77 thus obtained is in excellent agreement with the one togashi and kaneko observe and show in fig . 2 of their letter @xcite , when they see the system go through a transition for @xmath98 .
also , this verifies the other result of their simulations , shown in fig . 4 of their letter @xcite , where the rate of residence of the symmetry - broken state approaches zero , as @xmath99 . in our work on three - species systems
@xcite we have obtained similar results using van kampen expansion , and verified them in the diffusive ( fixation ) regime , using a kramers - moyal expansion , having imposed the condition that the system be near a critical transition , i.e. the smallest eigenvalue of the fokker - planck equation be zero .
the agreement between the two is excellent .
it is important to know the behaviour of the system , when it is closed , i.e. no migrations into and out of the container are possible , since the final state in the fixation ( broken - symmetry ) regime is the same
. the rate equations for this system , written for the concentrations of individual species , will be : @xmath100 and similarly for the other species .
if we consider the case when @xmath101 or @xmath102 with equal probability , the right - hand side of the rate equations will be identically zero .
one can directly find the steady - state solutions of the rate equations ( [ eq : rateq ] ) , i.e. the concentrations that reduce the left - hand side to zero .
for any number of species in the system , the centre , where all the concentrations are equal , is a solution of the rate equations ( [ eq : rateq ] ) .
when three species are present , the other solutions are those for which one of the species is alive , and the others are extinct . in the case of four and five species ,
the solutions are of the form @xmath103 ( modulo 4 or 5 , depending on the number of species in the system ) with all the other concentrations equal to zero .
the system with six species , except for the centre , has two other kinds of sets of solutions : one with @xmath104 ( modulo 6 ) and all the others zero ( i.e. three species alive and three dead ) , and one set with @xmath105 ( modulo 6 ) with the rest of concentrations equal to zero ( i.e. two `` opposite '' species alive and four dead ) . for any numbers of species in the system ,
the product of concentrations is conserved , as well as the total number of individuals .
this means that any trajectory with the product of concentrations constant will be a neutrally stable one . in the case of the neutral systems ,
the right - hand side is identically zero , which means that any state is a stable one .
we simulated copies of the four- or more - species cyclic system , for different system sizes .
these simulations started with equal individual populations of @xmath4 , @xmath6 , @xmath5 , and @xmath7 ( i.e. the centre ) .
we generated times for the next possible reaction event with exponential distribution as @xmath106 , where the rate of the cyclic or migrations as in the master equation above ( [ eq : mast4 ] ) is substituted .
( here @xmath107 is a random variable with uniform distribution in @xmath108 $ ] .
this way we get poisson distribution for the events , i.e. really independent events @xcite . )
the reaction which occurs first is then picked and the system is updated accordingly .
the process is repeated for a large number of events .
we observed the symmetry - breaking phenomenon for any system size , no matter how large it is . , @xmath6 , @xmath5 , and @xmath7 species in the `` fixation '' regime ( here @xmath82 , and system size @xmath109 ) .
in this particular realization the @xmath110 symbiotic pair wins over the @xmath111 one . ] in fig .
[ fig : tsabcd ] we show the time series for the four - species cyclic system , for @xmath109 .
contrary to the prediction from the solutions to the rate equations , for the six - species systems , out of two thousand realisations of the system , we never obtained a final state with only two species present .
all the copies of the system we simulated ended up with three species present , as the time series shown in fig .
[ fig : tsabcdef ] . in all the time series ,
time is measured in units of the system size @xmath84 . ) .
all the copies result in a state in which three species ( rather than two ) survive . ]
when more species are present in the system , the picture gets more complicated .
however , the general feature is that the system experiences `` fluctuations - generated forces '' , which push the system towards the boundary .
there the history of the system ends , since the boundary is absorbing .
when the number of species is odd , the population sizes of individual species oscillate with an amplitude that drifts with time , until they reach the boundary , one after another . on the other hand , when the number of species is even , `` alliances '' do not take long to form , and the history of the system always ends with half of the species winning over the other half .
a typical time series for an eight species system is shown in fig .
[ fig : ts8sp ] .
the fate of the system remains similar throughout the `` fixation '' regime , i.e. when the diffusion rate @xmath112 . in that regime ,
the system exhibits diffusive behaviour toward the boundary , and the migrations in and out are not frequent enough to bring it back to the centre . ) .
the history of the system ends in a state in which four species survive . ]
the fluctuations push the neutral system towards the boundary , too , but the history of the neutral systems ends either with only one species present , or with disconnected species , in the sense that they have no reason to coexist in symbiosis , since they have no common enemies . on the other hand ,
the history of the cyclic systems ends with two or more species present .
the surviving species do not react with one - another directly , but have common `` enemies '' .
this is quite reminiscent of symbiosis ( an alga + a fungus = a lichen , or a tree and mushrooms live together and help each - other fight common enemies ) , and yet another example of competition giving rise to cooperation . as a final remark regarding the even - number cyclic system , it is useful to try and `` coarse - grain '' its dynamics into two species : odd- and even - numbered .
for simplicity , let us consider the four - species system . in this case
: @xmath113@xmath114 and the master equation will transform as follows : @xmath115a_{odd } a_{even } + \right.\]]@xmath116 \right \}p(a_{odd},a_{even},t)\ ] ] this is exactly the master equation for kimura s two - species neutral drift system @xcite , if we recall that the odd even system size is @xmath117 .
duty has shown that this system is equivalent to a branching process @xcite with selection coefficient @xmath118 @xcite .
it is critical , and the extinction times scale as @xmath50 , and the probability of survival decays with time as @xmath119 , as verified from simulations ( see fig . [
fig : kimura ] ) . and
it maps onto the two - species neutral drift .
the slope of the line is @xmath121 for @xmath122 , and @xmath123 for @xmath120 . ]
it is easy to see that for systems with more species ( but always an even number of them ) we will obtain the same `` coarse grained '' master equation .
now let us go back to the volterra equations ( [ eq : voln ] ) .
they can be generalised further , resulting with the following rate equations : @xmath124 where @xmath8 is the number of species in the system .
the diagonal elements in both sums correspond to a malthus - verhulst equation .
the first sum now contains off - diagonal elements , which correspond to mutations of individuals from species @xmath24 to species @xmath36 ( or the other way around , depending on the sign of the rate @xmath125 .
these mutation rates will be very small . ) hence , the @xmath125 s and @xmath126 have opposite signs ( and also @xmath35 and @xmath127 ) .
the terms of the form @xmath128 describe what happens when an @xmath24 individual runs into a @xmath36 individual : when @xmath129 the @xmath24 individual `` eats '' a @xmath36 individual and reproduces , and the other way around when @xmath130 .
the master equation for the generalised volterra system , written in terms of the `` shift '' operators ( [ eq : shift ] ) : @xmath131@xmath132 where the way the master equation is written imposes that the coefficients @xmath133 be chosen as equal to @xmath125 if @xmath134 , and zero otherwise .
similarly , @xmath135 if @xmath136 , and zero otherwise
. now we are ready to attempt the van kampen expansion of the above master equation .
as before , the variables @xmath137 are split into a non - fluctuating and a fluctuating part , as in eqs .
( [ eq : splitv ] ) , which transforms the probability density function similarly to ( [ eq : vartrans ] ) . after substituting everything into the master equation
, we obtain a fokker - planck equation as in ( [ eq : fpsurv ] ) , where the a and b matrices have dimensions @xmath138 .
the diagonal elements of the a - matrix will be given by @xmath139 , and the off - diagonal ones by @xmath140 .
the existence and stability of the solutions to the volterra equations ( [ eq : voln ] ) has been object of extensive studies .
when the system is stable , the a - matrix will have negative eigenvalues , which , using the equation ( [ eq : avgfl ] ) , will mean that the average value of the fluctuations will decay to zero . here
we are in luck , since , after a careful inspection , we can see that its elements are those of the jacobian @xmath141 calculated at the fixed point .
if the eigenvalues of this jacobian are negative , the fixed point is stable .
the stability of the solution can be determined from the _ routh - hurwitz criteria_. the interested reader can find a treatment of these and further references in chapter 6 of the book by pielou @xcite .
however , the application of routh hurwitz criteria can be tedious and cumbersome , especially when dealing with large systems .
the technique of _ qualitative stability _ helps enormously .
it was initiated by the economists quirk and rupert @xcite , and applied to ecological systems by may @xcite , levins @xcite , and jeffries @xcite .
see chapter 6 of the book by edelstein
keshet @xcite for an extensive review of this subject matter .
the b - matrix has diagonal elements of the form @xmath142 and off - diagonal ones : @xmath143 .
they are linear in coefficients @xmath133 and @xmath144 , which means that the equations for the second moments of the fluctuations ( obtained from their general expression ( [ eq : varflu ] ) ) will also be linear in the rate coefficients .
this will lead to a size - induced transition as the one calculated for the autocatalytic loops , and observed in simulations by us for the three - species systems @xcite , and togashi and kaneko for four - species systems @xcite .
the transition observed by togashi and kaneko @xcite in the four - species autocatalytic loops is not specific to systems with a very small number of particles / individuals .
it is the same critical transition we have previously @xcite observed for three - species systems .
this transition corresponds to a crossover from a fully symmetric ( `` neutral '' ) state in the high - migration regime , to a `` fixation '' state for low - migration rates , in which the symmetry is broken in favour of one or more species .
the `` fixation '' regime in the four - species system exhibits a symbiosis effect , when species @xmath4 and @xmath5 join their efforts against species @xmath6 and @xmath7 , and the final state of the system is one in which one of the pairs has completely `` eaten up '' the other .
this transition is present for any finite system size . in the high - migration regime
the system allows for a linear noise approximation , exhibiting itself as an ornstein - uhlenbeck process .
since the system size is always finite ( no matter how large it is ) , there is a value of the migration probability per particle ( or diffusion rate ) for which the fluctuations of the concentrations become of order one , and the system undergoes a critical transition . the critical diffusion rate varies with system size as @xmath145 , and the product @xmath146 , i.e. the number of migrants per unit time , necessary for the symmetry to be preserved in the system ( all species to survive ) is of the order 1 .
this result is a bit counterintuitive , since it does not depend on system size .
the analytical calculations are in excellent agreement with the simulation results , obtained for three- and four - species systems @xcite .
those analytical calculations suggest that all the loop - like autocatalytic systems will exhibit the same critical transition .
the form of the equations for the moments of the fluctuations suggests that the generalised lotka - volterra systems will also exhibit a similar transition .
this situation , known as diffusion - limited reaction , has manifested itself and been observed in physical systems low dimensions , when diffusion is not efficient in mixing the reactants .
a physical example is the ovchinnikov - zeldovich segregation phenomenon @xcite .
in such a situation the hope is that sufficient migration between habitat patches will save the system from extinction .
abta and shnerb @xcite show that the systems in which only the predator is allowed to migrate ( such as herbivore plant or parasite insect plant systems , like the prickly pear cactus and the moth _
cactoblastis cactorum _ in eastern australia ) may support oscillations in noisy environment .
much work is yet to be done in the future . | we discuss an autocatalytic reaction system : the cyclic competition model @xmath0 , @xmath1 , @xmath2 , @xmath3 ) , as well as its neutral counterpart .
migrations are introduced into the model . when stochastic phenomena are taken into account , a phase transition between a `` fixation '' and a `` neutral '' regime is observed . in the `` fixation '' regime , species @xmath4 and @xmath5 form an alliance against species @xmath6 and @xmath7 , and the final state is one in which one of the symbiotic pairs has won .
the odd even `` coarse grained '' systems is mapped onto the two species neutral ( kimura ) model . in the `` neutral '' regime ,
all four species survive for long ( evolutionary ) times .
the analytical results are checked against computer simulations of the model .
the model is generalized for @xmath8 species .
also , a generalized version of the volterra model is analysed . |
the antennae ( ngc 4038/39 = arp 244 ) is a pair of late type spirals in the course of merging .
it is the youngest in toomre s ( 1977 ) dynamical age sequence of 11 interacting and merging systems and comfortably nearby .
this makes it a well - studied system all over the wavelength range from x - ray through radio wavelengths .
the interaction has triggered a strong burst of star formation that has also produced a large number of @xmath1 bright star clusters as first observed with hst by whitmore & schweizer ( 1995 ) ( * ws95 * ) .
many of those bright clusters seem to be young globular clusters ( * gcs * ) due to their small effective radii and high luminosities .
the luminosity function ( * lf * ) of this bright star cluster system looks like a power law @xmath2 with no hint of a turnover at fainter magnitudes down to the completeness limit of @xmath3 mag which , at the distance of the antennae ( 19.2 mpc for h@xmath4 ) , corresponds to @xmath5 mag ( ws95 ) . in fritze v. alvensleben ( 1998 ) ( hereafter * pap.i * ) i analysed ws95 s hst data with evolutionary synthesis models for single burst single metallicity populations ( * ssps * ) using stellar evolutionary tracks for various metallicities from the geneva group and a scalo imf from 0.15 to 60 m@xmath6 ( fritze v. alvensleben & burkert 1995 , * fb95 * ) . in analogy to the ysc system in ngc 7252 which , like the antennae , is an sc sc merger , we assume that the metallicity of the ysc population that forms out of the spirals ism is @xmath7 ( fritze v. alvensleben & gerhard 1994 ) . in pap.i individual ages are determined from ( @xmath8 ) colours for the 550 star clusters with i band detections .
the age distribution clearly reveals two populations of clusters , @xmath9 old gcs from the original spirals , and @xmath10 * young * star clusters ( * yscs * ) with ages in the range @xmath11 yr formed in the ongoing interaction - induced starburst .
only the secondary population of yscs will be considered in the following .
meurer ( 1995 ) was the first to point out the possible importance of age spread effects on the future evolution of the lf of a ysc system . with individual ysc ages and the fading in various passbands as given by our ssp models
we were able to calculate the future evolution over a hubble time of the lf and of the colour distribution of the ysc system under the unrealistic assumption that all clusters will survive . at an age of @xmath12 gyr , when age differences of the order of @xmath13 yr do no longer play any role ,
the lf is shown to be undistinguishable from a typical gc lf ( i.e. gaussian with a turnover at @xmath14 mag appropriate for the metallicity [ fe / h ] @xmath15 ( ashman et al .
1995 ) and @xmath16 mag ) .
the colour distribution will by then also be compatible with the one observed on old gc systems .
while accounting for stellar mass loss , this modelling , however , did not take into account dynamical effects like two - body relaxation , dynamical friction , or disk shocking that might act heavily on a ysc system over a hubble time .
already for the milky way , where the potential is rather well known , it is difficult and not uncontroversial to model the dynamical processes on individual clusters ( e.g. chernoff & weinberg 1990 , fukushige & heggie 1995 , gnedin & ostriker 1997 , vesperini 1997 ) . this is , of course , even more difficult , if not impossible , in an ongoing merger like the antennae . before one can try and quantify the efficiency of various destruction mechanisms , the intrinsic mf underlying the presently observed lf has to be determined .
knowing individual cluster ages offers the possibility to use m / l values from evolutionary synthesis models to derive the mass function ( * mf * ) underlying the presently observed lf and this is what we attempt in this letter .
since the age distribution of the yscs is strongly peaked within @xmath17 yr and the yscs have a mean distance to the galaxy center of @xmath18 kpc , a ysc on average can not have had more than 1 or 2 revolutions .
we do not expect the mf therefore to already be significantly affected by cluster destruction processes .
rather we expect the presently derived mf to reflect the mf produced by the cluster formation process .
the mass spectra of molecular clouds , molecular cloud cores , open clusters , and the lf of giant hii regions ( in non - interacting galaxies ) all are power laws with exponents @xmath19 in the range @xmath20 ( e.g. solomon et al .
1987 , lada et al .
1991 , kennicutt 1989 , see harris & pudritz 1994 and elmegreen & efremov 1997 for overviews ) as is the mass spectrum of open clusters in the milky way and the lmc ( e.g. van den bergh & lafontaine 1984 , elson & fall 1985 ) .
both the lf and the mf of old gc systems are gaussians with typical parameters @xmath21 mag , @xmath22 mag , and @xmath23 , @xmath24 , respectively ( e.g. ashman et al . 1995 ) .
the question immediately arises : is the transition from a power law molecular cloud mass spectrum to a gaussian old gc mass spectrum performed in the star / cluster formation process or by secular destruction effects within a gc system ? or , else , is already the mass spectrum of molecular clouds or molecular cloud cores different in strongly interacting and starbursting galaxies from what it is in normal spirals ?
on the basis of individual ysc ages we use our ssp models giving @xmath25 in the passbands @xmath26 ubvrik as a function of time to derive masses of individual clusters from their observed v luminosities .
this is done for all the 393 yscs with ages @xmath27 yr and v luminosities brighter than the completeness limit @xmath5 mag .
it is stressed that our model @xmath28 - values include the decrease in cluster mass due to stellar mass loss ( cf .
pap.i ) , but not that due to the evaporation of stars from the cluster .
the mf we recover in this way from the presently observed lf is presented in fig.1 .
a gaussian with @xmath29 and @xmath30 , normalised to the number of yscs in the histogram , is overplotted .
the intrinsic mf we obtain for the yscs brighter than the completeness limit in the antennae clearly looks log - normal in shape with the maximum at a mean ysc mass of @xmath31 .
stellar mass loss within the clusters from the present mean age of @xmath32 yr through an age of @xmath12 gyr will lead to a decrease in mass of @xmath33 for a scalo imf ( @xmath34 for salpeter ) . and
@xmath30 is overplotted , normalised to the number of yscs in the histogram . ]
thus , without any destruction or evaporation effects the mean mass of the secondary gc system in the antennae at the age of a hubble time would be @xmath35 . a cluster with this mean mass would have @xmath36 mag at an age of 12 gyr according to our models .
this is the position of the maximum of the gaussian ysc lf we obtain at a hypothetic ysc age of 12 gyr ( cf .
6a in pap.i ) .
the agreement is no surprise since this is , in fact , the way how we obtained the parameters for the gaussian in fig.1 .
we stress that these parameters are derived from the lf in pap.i and are not the result of any fit to the cluster mf in fig.1 .
remarkably enough , the parameters of this gaussian which in fig.1 is seen to reasonably describe the mf of the secondary gc population are quite similar to those given by ashman et al .
( 1995 ) for the milky way and m31 gc systems . using evolutionary synthesis results from worthey ( 1994 ) for @xmath37 ,
ashman et al .
find @xmath38 and @xmath39 for the milky way and @xmath40 and @xmath41 for m31 .
the mf in fig .
1 thus seems compatible with the bulk of the yscs really being young gcs rather than open clusters or associations , as was already indicated by their small effective radii and high luminosities ( ws95 ) . from the model side ,
uncertainties in the determination of ysc ages ( and hence masses ) on the basis of their @xmath42 colours are dominated by uncertainties in the ysc metallicities .
age uncertainties due to metallicity uncertainty ( @xmath43 ) are estimated to be of the order of @xmath44 .
the uncertainty in m / l at @xmath45 due to the age uncertainty is @xmath46 % and the uncertainty in m / l at all ages @xmath47 yr due to the metallicity uncertainty is @xmath48 % .
this leads to an overall uncertainty in the m / l of @xmath49 % .
measurement uncertainties are @xmath50 mag for the observed @xmath42 colours and @xmath51 mag for v magnitudes for yscs brighter than the completeness limit .
the uncertainties that the inhomogeneous reddening over the body of ngc 4038/39 brings along for the derived cluster ages and masses are difficult to quantify .
only a global average value is given for the internal reddening of the ysc population in ngc 4038/39 ( @xmath52 mag ( ws95 ) ) and applied before our age - dating , but dust lanes and unrelaxed structures are seen all over ngc 4038/39 . the very good agreement ( @xmath53 yr ) between ages determined from @xmath54 and @xmath42 colours ,
however , indicates that for the bright clusters seen on the short u exposure the reddening does * not * seem to significantly deviate from the average reddening we use .
it should be noted in this context that inclusion of yscs fainter than the completeness limit which tend to have significantly larger observational uncertainties does * not * affect the agreement with the gaussian in fig.1 or its parameters ( beyond the normalisation to the number of clusters in the histogram ) .
* we conclude that the secondary gc population in the antennae is formed with a log - normal mass distribution very similar to the one in the milky way or m31 .
it is not necessarily the secular evolution but rather the cluster formation process that produces the gaussian mass spectrum observed on old gc systems .
* since uncertainties in the mf from observational errors can not be calculated in any straightforward way in the analysis presented here , we are currently trying an independent approach .
we draw yscs at random from different intrinsic mfs ( gaussians , power - laws ) , randomly assign ages to them from different age distributions ( clusters formed uniformly over the burst duration or at an increasing or decreasing rate ) , and calculate their present luminosities and colours to which , then , observational errors can be added , again at random from the observed luminosity - dependent error distribution .
comparison of the resulting model lfs and colour distributions with the observed ones should then allow to constrain the intrinsic mf ( kurth et al . , _ in prep . _ ) .
a somewhat similar procedure is used for yscs in ngc 7252 and ngc 3921 by miller & fall ( 1997 ) , who find that power - laws are preferred over gaussians for the mfs of these ysc systems .
ngc 7252 and ngc 3921 are the two oldest merger remnants from toomre s ( 1977 ) list .
large enough ysc systems have been detected in both of them to define the bright end of their lfs ( whitmore et al .
1993 , miller et al .
1997 , schweizer et al . 1996 ) .
distances , however , are larger than to ngc 4038/39 by factors 3 and 4 for ngc 7252 and ngc 3921 , respectively , pushing the completeness limit to significantly higher luminosities .
the higher mean ages of the yscs in these cases ( @xmath55 myr for ngc 7252 and @xmath56 myr for ngc 3921 , depending on metallicity ) add another argument in favour of them being young gcs , since they have survived @xmath57 ( cf .
schweizer et al . 1996 ) .
the lf we * calculate * for the yscs in the antennae at an age of @xmath58 gyr does show a marginal turnover at the expected @xmath59 mag for yscs brighter than the completeness limit , indicating that by this age the distortion of the lf with respect to the mf due to age spread effects from the finite ysc formation timespan ( of the order @xmath60 myr ) is already less important . in order to estimate
if a turn - over in their ysc lfs could be expected in ngc 7252 and ngc 3921 , we assume that their ysc systems have the same mf as the ysc system in the antennae .
then , we can calculate the mean absolute magnitude @xmath61 from @xmath62 at the above - quoted mean ages .
we obtain @xmath63 mag and @xmath64 mag for yscs in ngc 7252 and in ngc 3921 , respectively .
these luminosities are close to the 90 % completeness limiting magnitudes of @xmath65 ( pc ) and @xmath66 ( wf ) for ngc 7252 ( miller et al .
1997 ) and of @xmath67 for ngc 3921 ( schweizer et al .
1996 ) . together with the difficulty of disentangling the old gc population from the yscs at these advanced ages the fact that no turnover is detected in the lfs for ngc 7252 and ngc 3921 does not seem to rule out a gaussian mf similar to the one we obtain in the antennae
quantitatively , nothing is known about the external dynamical effects on gc systems in interacting and merging galaxies . for the milky way , where dynamical modelling on the basis of the galactic potential is possible and has been done by many groups ( e.g. chernoff & weinberg 1990 , gnedin & ostriker 1997 , fukushige & heggie 1995 , vesperini & heggie 1997 , vesperini 1997 , 1998 ) it is clear that the gc system observed today
is only `` the hardiest survivors of a larger original population '' ( harris 1991 ) .
vesperini & heggie ( 1997 ) present n - body simulations including effects of stellar evolution , two - body relaxation , disk shocking , and the tidal field of the milky way . studying the secular evolution of a number of gc systems with different initial mfs , vesperini ( 1998 )
shows that * if the initial mf is log - normal then this log - normal shape and its parameters are conserved over a hubble time * despite the destruction of a large part ( @xmath68 ) of the original gc population .
while evaporation and dynamical friction preferentially destroy low and high mass clusters , respectively , both processes balance each other in the case of a log - normal ( @xmath69 equilibrium ) initial gc mf , so that no net selective destruction of specific gc masses results .
if the gc destruction processes were similar in the antennae and in the milky way , and if , as indicated in fig.1 , the initial gc mf really were close to vesperini s equilibrium mf , then the gaussian mf could survive a hubble time with its parameters @xmath70 and @xmath71 essentially unchanged despite the likely destruction of a large fraction of the ysc system seen today . it will be interesting to analyse more ysc systems in the way we did in order to see if and in how far the ysc mf is universal or depends on parameters of the progenitor galaxies or the interaction event .
moreover , studying secondary gc systems on an age sequence of interacting , merging and merger remnant galaxies should allow to directly `` observe '' both the time evolution of the lf and of the mf and thus offer a unique possibility to study the effects of dynamical processes _ in situ_. the turnover of an old gc population at @xmath72 mag occurs around @xmath73 at virgo cluster distances and is well within the reach of 10 m telescopes .
hst imaging allows to identify clusters in regions not too crowded for ground - based mos .
the uncertainties in the metallicities and more important in the individual reddening of yscs in the antennae can be substantially reduced with spectroscopic observations of ysc selected from the hst images . with mos facilities on 10 m class telescopes
this should be possible in the nearest future .
one question that poses itself immediately in the context of the cluster formation mechanism is whether the molecular cloud mass spectrum in a strongly interacting and starbursting galaxy like the antennae does really have the same power - law form as that in an undisturbed spiral forming stars at a level orders of magnitude lower ?
elmegreen & efremov ( 1997 ) pointed out that the high pressure of the ambient ism produced in a strong interaction might favor the production of more massive clouds .
information about the molecular cloud mass spectrum in merger - induced strong starbursting systems seems a prerequisite for the study of the star and cluster formation processes .
first spectroscopic observations of a handful of yscs in ngc 7252 ( schweizer & seitzer 1993 , 1998 ) confirms the metallicity of @xmath74 ( cf .
fb95 ) we predicted on the basis of the progenitor galaxies ism abundances .
this enhanced metallicity ( with respect to the primary gc population ) raises the question in how far the secondary gc formation process is comparable to the primary one in the early universe ?
the mf of the yscs in the antennae seems to be log - normal with parameters @xmath70 and @xmath71 very similar to those of the milky way gc system .
this suggests that the cluster formation process and not the dynamical evolution produce the gaussian mf .
the close agreement of the ysc mf we obtain with vesperini s equilibrium initial gc mf seems to indicate that the shape and parameters of this mf may survive a hubble time despite destruction of a large fraction of today s yscs . in the older merger remnants
ngc 7252 and ngc 3921 the completeness limit is close to the turn - over luminosity expected in case their mfs were similar to the one in the antennae .
ashman , k. m. , conti , a. , zepf , s. e. , 1995 , aj , 110 , 1164 chernoff , d. f. , weinberg , m. d. , 1990 , apj , 351 , 121 elmegreen , b. g. , efremov , y. n. , 1997 , apj , 480 , 235 elson , r. a. , fall , s. m. , 1985 , pasp , 97 , 692 fritze
v. alvensleben , u. , 1998 , a&a 336 , 93 ( * pap.i * ) fritze v. alvensleben , u. , gerhard , o.e . , 1994
, a&a 285 , 751 fritze
v. alvensleben , u. , burkert , a. , 1995 , a&a 300 , 58 ( * fb95 * ) fukushige , t. , heggie , d. c. , 1995 , mn , 276 , 206 gnedin , o. y. , ostriker , j. p. , 1997
, apj , 474 , 223 harris , w.e .
1991 , araa , 29 , 543 harris , w.e . , pudritz , r. e. , 1994 , apj , 429 , 177 kennicutt , r. c. , 1989 , apj , 344 , 685 lada , e. , bally , j. , stark , a. a. , 1991 , apj , 368,432 meurer , g. r. , 1995 , nat , 375 , 742 miller , b. w. , fall , s. m. , 1997 , aas , 119 , # 115.04 miller , b. w. , whitmore , b. c. , schweizer , f. , fall , s. m. , 1997 , aj , 114 , 2381 schweizer , f. , seitzer , p. , 1993
, apj 417 , l29 schweizer , f. , seitzer , p. , 1998
, aj 116 , 2206 schweizer , f. , miller , b. w. , whitmore , b. c. , fall , s. m. , 1996 , aj , 112 , 1839 solomon , p. m. , rivolo , a. r. , barrett , j. , yahil , a. , 1987 , apj , 319 , 730 toomre , a. , 1977 , in _ the evolution of galaxies & stellar populations _ , eds . b. m. tinsley , r. b. larson , yale univ .
obs . , new haven , p. 401
van den bergh , s. , lafontaine , a. , 1984 , aj , 89 , 1822 vesperini , e. , 1997 , mn , 287 , 915 vesperini , e. , 1998 , mn , 299 , 1019 vesperini , e. , heggie , d. c. , 1997 , mn , 289 , 898 whitmore , b.c . ,
schweizer , f. , 1995 , aj 109 , 960 ( * ws95 * ) whitmore , b.c . ,
schweizer , f. , leitherer , c. , borne , k. , robert , c. , 1993 , aj 106 , 1354 worthey , g. , apjs , 95 , 107 | the antennae is a pair of late type spirals in the course of merging .
the interaction triggered an ongoing strong burst of star formation that also produced a large number of young star clusters ( yscs ) , many of which seem to be young globular clusters ( gcs ) .
the observed luminosity function of the ysc system is a power - law .
we use evolutionary synthesis models for star clusters in comparison with hst wfpc observations of the ysc system in the antennae to analyse the mass function of this @xmath0 yr young star cluster system . properly accounting for age spread effects by individually age - dating the yscs we find that the intrinsic mass function ( mf ) is log - normal in shape and in intriguing agreement with the mf of old gc systems .
we discuss this mf in the context of cluster formation and dynamical effects in the tidal field of the parent galaxy and speculate about its future evolution . |
the role of electronic interactions on the observed properties of semiconductor quantum dots has increasingly been found to be of vital importance , as many of the papers in these proceedings and elsewhere illustrate . more and more ,
transport and capacitance experiments , as well as detailed studies of far - infrared response , require the consideration of interactions in order to understand the observed experimental features .
since a number of experiments explore specifically the electronic level spectrum , it is of interest to contrast these studies with a detailed theoretical analysis of the role of interactions .
we present here an analysis of the energy level statistics of a quantum dot system as function of variable interaction strength , and then as function of an applied magnetic field . as the level spectrum is considered in detail , the correspondence with the dynamical integrability of a classical system is also of interest . it is anticipated that as the coulomb interaction is introduced , the dynamics would in general become chaotic and this would in turn be reflected in various statistical properties of the level spectrum .
it has been known for some time now that as a classical system becomes non - integrable @xcite , the corresponding quantum system exhibits a level spacing distribution given by the ` wigner surmise ' derived in the context of random matrix theory ( rmt ) @xcite .
in fact , this behavior has been verified in a number of theoretical and experimental systems , although typically the classical non - integrability is due to disorder or boundary ( geometrical ) effects .
examples discussed in the literature include small disordered metallic particles @xcite , particles moving in a variety of ` stadia ' @xcite , and in two - dimensional antidot arrays @xcite .
rmt also has been used to make definite predictions on the statistical distribution of coulomb blockade peak amplitudes @xcite .
this behavior has in fact been shown recently to be a good description of quantum dot systems in beautiful experimental realizations @xcite .
one should also mention that recent work on interacting systems , some without an obvious classical counterpart , has also shown that these exhibit the level repulsion characteristic of quantum versions of non - integrable classical systems @xcite .
moreover , recent analysis of the level spectrum of excitons in quantum wells ( via photoluminescence excitation spectroscopy ) has shown evidence of level repulsion @xcite . a level structure described by rmt has been sought recently in the charging ( or addition ) spectrum of quantum dots @xcite
. detailed analysis of experimental charging energies , after proper subtraction , would be expected to yield the single - particle like _ excitation _ spectrum described by rmt .
unfortunately , it appears that the extraction of this excitation spectrum is obscured by the systematic shift in the charging energies , and the resulting level spacing is described by a normal distribution , rather than by the rmt functions .
this would then suggest that a more direct measurement of the excitation spectrum ( via perhaps nonlinear tunneling currents ) would be desirable ( although clearly difficult experimentally beyond the first few excitations @xcite ) .
one should also mention here that it is believed that the nonlinear transport experiments explore mainly the excitations of the center of mass of the system ( in the typically parabolic quantum dots ) @xcite , due to the strong electronic correlations suppressing most of the tunneling ` channels ' .
this prevalence of the center of mass excitations is however expected theoretically to diminish as the energy of the excitation increases @xcite .
this regime would be reached only as the bias voltage is raised in transport experiments , and makes it then difficult to achieve in practice .
we hope , however , that the results presented here would motivate more experimental work in this direction .
we have studied the effects of particle interaction on the classical integrability of a system of two masses moving inside a @xmath0-dimensional billiard .
we find , in general , that the motion is strongly chaotic ( typically exhibiting ` soft ' chaos , with a mixed dynamics , where regions of phase space are still periodic or quasi - periodic ) , and with a strong dependence on the characteristic interaction length and strength . as perhaps one of the simplest examples ( see ref .
@xcite for a description of the general case ) , consider two particles of equal masses moving in a 1d box defined in the interval @xmath1 ( we measure all lengths in terms of the box size ) .
the particles are assumed to interact via a screened potential @xmath2 , where @xmath3 is the inverse screening length .
notice that this potential goes to a hard - core @xmath4function when @xmath5 , and the particles behave then as non - interacting but impenetrable points . in that case
, the dynamics can be integrable in special cases of the mass ratios @xcite . in this sense
, @xmath3 plays the role of a perturbation parameter which changes the degree of integrability of the system , as it determines the effective ` radius ' of the particles for a given total energy .
direct calculations show that this is indeed the case .
the problem could in principle be solved by direct integration of the equations of motion , but we find convenient to transform it to a set of center - of - mass and relative coordinates @xmath6 , and @xmath7 , respectively , where the total mass @xmath8 , and the reduced mass @xmath9 .
these equations define a two - dimensional space of coordinates @xmath10 . in this space
, we have a new set of equations for the boundary of the billiard , say @xmath11 .
the hamilton equations are transformed to @xmath12 , @xmath13 , and @xmath14 \nonumber \\
\dot { p } & = & \sum _ { j } { b}_j({p , p } ) \ , \delta[f_j({r , r } ) ] .
\end{aligned}\ ] ] the functions @xmath15 and @xmath16 describe the change in the momenta @xmath17 and @xmath18 , due to the bounce on the @xmath19-th wall .
notice that these equations describe the motion of a single ` hyperparticle ' in the two - dimensional @xmath20-space .
this _ hyperbilliard _
description can be generalized to any number of dimensions @xcite .
notice that bounces of the hyperparticle in the hyperbilliard correspond to bounces of the masses in the real / original dot .
the walls of the billiard cause the breaking of translational symmetry of the system , and as a consequence , the center - of - mass ( cm ) momentum is no longer a constant of motion . in the case of non - interacting and equal - mass particles , the changes in the cm momentum are determined only by the geometry of the billiard . in our case , however , the interaction couples the cm and relative momenta after each bounce , which in turn depend on the momenta of each of the original masses .
* dynamical map*. we should also mention that these equations in @xmath20-space provide an interesting description which is also extremely useful : in between bounces , the hyper - particle moves freely along the @xmath21-axis , whereas the interaction acts only along @xmath22 .
these two motions are independent , and only become correlated at each bounce , when the different momentum components are changed , while keeping the total energy constant . understanding this fact allows one to describe the motion in terms of a _ dynamical map _ connecting the different bounces .
if the coordinate of the hyperparticle is @xmath23 at the time of the @xmath24-th bounce , then the time spent until the next bounce on the @xmath19-th wall is obviously the same along _ both _ components , and one can then write @xmath25 . since the cm motion is that of a free particle , except for the collisions with the walls , @xmath26 can be calculated simply .
similarly , for a pure coulomb potential ( @xmath27 ) , @xmath28 can be calculated analytically , so that the previous equation can be written in a more explicit form , @xmath29 , where @xmath30 is now the time elapsed going from @xmath31 to @xmath32 , expressed in terms of the time @xmath33 spent by the particle from the turning point to @xmath20 , @xmath34 . here
@xmath35 , represents the energy left for the relative motion , as @xmath36 is the total energy of the two - mass system . for a weakly screened potential , @xmath37 , we can expand @xmath38 to first order and obtain a similar expression , where @xmath36 is scaled to @xmath39 .
what follows , after these definitions is to characterize all the different possible trajectories in the triangular region in @xmath20-space . a simple algorithm
can then be obtained to determine the poincar surfaces of section .
this nontrivial ( and clearly nonlinear ) algebraic map provides then a full description of the dynamics .
its use ( in lieu of the direct integration of the equations of motion ) simplifies calculations a great deal , and allows one to better characterize the system , as we describe below .
* poincar sections*. in order to characterize the motion of the two interacting particles ( or the hyperparticle with its many degrees of freedom ) , we explore poincar sections of the resulting phase space . having a four - dimensional space in this case @xmath40 , we select to show those where one of the particles is at one end of the box ( notice that the energy is a constant of motion here ) . figure 1 shows a typical poincar section for @xmath41 and @xmath42 ,
obtained using the map described above ( and which is virtually identical to the one obtained directly from integration of the equations of motion @xcite ) .
notice that chaotic trajectories nearly fill the available phase space ( for this given total energy ) .
increasing values of @xmath3 give chaotic orbits that fill more of the available phase space .
moreover , there are also a number of islands of stability , as expected from the kam theorem @xcite , near the fixed point corresponding to periodic symmetric motion in the non - interacting system ( and indicated in the figure with a cross on the right axis ) .
in fact , for all values of @xmath36 and @xmath3 , the initial condition where @xmath43 , and @xmath44 gives rise to a periodic orbit .
other islands also appear purely due to the interaction , as can be seen near @xmath45 , and are associated with a sort of correlated motion of the two masses .
general values of @xmath36 and @xmath3 give rise to this type of mixed dynamics , with chaotic and regular trajectories sharing the available phase space .
this increasing degree of ` soft chaos ' in the system ( as @xmath3 increases , for example ) , should be reflected in the level statistics of the corresponding quantum mechanical system , as we describe explicitly elsewhere in these proceedings .
one naturally expects that this effect of interactions turning a regular system into a non - integrable one would be rather pervasive , regardless of the type of particle confinement and details of the interactions . in the following section we illustrate this effect in a somewhat different quantum mechanical model of a quantum dot .
here , we will use a model of a quantum dot which has been very successful in the description of recent experiments in these structures @xcite .
the quantum dot is modeled as a parabolic potential well with circular symmetry , and the frequency ( or curvature ) is chosen to fit characteristic single - particle excitation energies in these devices . in this case , the few - particle problem , which includes fully the effect of interactions , can be solved quite accurately ( numerically ` exactly ' ) even in the presence of magnetic fields . the approach is based on a canonical ( jacobi ) transformation to a set of auxiliary harmonic oscillator generating operators which allow one to write the interaction matrix elements in a closed analytical form , easily calculable @xcite .
this , together with the ability to separate the center of mass ( cm ) degree of freedom from the ` relative ' ones ( a feature almost exclusive to the parabolic confinement potential ) , allows one to completely characterize and solve for the spectrum of up to three electrons in this parabolic well . correspondingly , the spectrum can be characterized by @xmath46 , where the first term gives the cm manifold , and @xmath47 is obtained from the relative - motion part of the hamiltonian . in the case of three particles ,
this latter part has diagonal elements given by @xmath48 , which are mixed by the coulomb interaction ( taken here as given by @xmath49 , with @xmath50 a background dielectric constant which would be assumed variable in the results section below ) .
notice that in all these expressions , @xmath51 arise from the auxiliary harmonic oscillators introduced in the canonical transformation @xcite , with @xmath52 /2 $ ] .
here , @xmath53 characterizes the single - particle harmonic confinement potential of the quantum dot ( with a typical value of @xmath54 12 mev ) , and @xmath55 is the cyclotron frequency of the electron .
due to the rotational symmetry of the system , the coulomb interaction only couples states with the same relative angular momentum , given by @xmath56 .
this is important , for apart from making the calculation simpler , it also allows us to carry out matrix diagonalizations with extremely high accuracy and for very many eigenstates ( typically one to two thousand levels ) .
this complete convergence is obviously important if one is interested in the analysis of the level spacings , as we now proceed to do .
* level statistics results*. the statistical analysis of the levels obtained as described above for three particles can be carried out according to the typical prescriptions in the literature .
the level spectrum is used to extract a slowly - varying density of states with a smooth energy dependence , characteristic of the system at hand .
this _ unfolding _ procedure then leaves one to study the structure of the level fluctuations of the spectrum , on which a number of statistical tests and statements can be made @xcite .
the unfolding here is performed by defining the ` staircase ' function @xmath57 which gives the cumulative number of states below @xmath36 , and then fitting this to a smooth polynomial ( typically of fourth degree ) , @xmath58 .
the ` linearized ' or unfolded level sequence is then obtained from @xmath59 , where @xmath60 is the original sequence .
this process , although not unique , gives similar results to other unfolding procedures ( see @xcite for a good discussion ) .
one can perform a number of statistical analyses . here
we focus on the nearest - neighbor spacings ( nns ) , obtained from @xmath61 , and calculate the probability density function @xmath62 for a given sequence .
it is useful to define the integrated probability function @xcite , @xmath63 .
notice that @xmath64 is nothing but the total number of nns below a given @xmath65 , and can be uniquely calculated , without any dependence on the specific binning used to calculate the typical histogram representations of @xmath62 . in
what follows , we use @xmath64 to make quantitative statements , but revert to showing the more conventional @xmath62 . figure 2
shows a sequence of @xmath57 staircase curves for different values of the background dielectric constant @xmath50 , defined above .
the sequences analyzed have all the same value of relative angular momentum , @xmath66 , and total spin @xmath67 , and for a magnetic field value of @xmath68 t .
increasing values of @xmath50 would produce a progressively weaker value of the coulomb interaction , and vice versa .
we have varied @xmath50 such that the interaction varies by up to a factor of 50 . in the figure , the curve labeled 1 corresponds to the value of @xmath69 , found in gaas , where most quantum dots are defined . increasing the effective coulomb interaction ( with @xmath70 ) , produces the smoothest staircase function ( labeled 10 ) , completely devoid of the harmonic oscillator ` steps ' in weaker interactions ( as seen here for relative interaction strength 1/5 and 1/3 ) .
one expects that for interaction 10 , the level mixing produced would be substantial , resulting in a level structure well described by the rmt distributions , as seen in other systems .
we find that this is not the case here .
in fact , if one analyzes the appropriate histograms for the nns distribution function @xmath62 , the interaction strength shows its effect quite clearly , as shown in fig .
, we show the corresponding @xmath62 for three different interaction strength ( or @xmath50 ) values .
the anticipated poisson distribution function one obtains for a ` generic ' integrable system @xcite is similar to the case for interaction 1/3 .
however , notice that this @xmath18 here goes to zero even _
faster _ than the expected poisson form , reflecting the non - generic character of the delta - function like distribution of the pure harmonic oscillator system .
moreover , as the interaction becomes stronger , notice that the distribution functions have a maximum not at zero @xmath65 , but rather at a finite value .
this behavior is more in agreement with the rmt predictions , where @xmath18 would be expected to be given by a gue function ( given the finite magnetic field ) , where @xmath71 .
one should notice , however , that in all cases we have studied , @xmath62 never fully reaches the anticipated gue for a fully chaotic system .
the reason for this lack of full crossover into the gue distribution is perhaps associated with the mixed dynamics of the corresponding classical system . in such cases , it has been argued that the level distribution can be seen as a superposition of poisson and gue functions , with weights corresponding to the coverage in phase space for the integrable and non - integrable regions , respectively @xcite . at this point , however , we have not analyzed the classical system that directly models this quantum dot , and expect to report on this relation elsewhere .
however , it may also be the case that this incomplete crossover to a gue arises from the peculiar non - generic non - poissonian distribution function of the integrable ( non - interacting ) system composed by overlapping harmonic oscillators , and/or remnant hidden symmetries .
we have also fitted the @xmath72-functions to the well known brody distribution , @xmath73 , providing a ( phenomenological ) measure of the crossover ( @xmath74 for poisson , @xmath75 for goe ) . in fig . 2 and for unit interaction strength ,
we get @xmath76 , while @xmath77 for interaction 10 .
furthermore , the non - typical character of the level distribution in the harmonic oscillator is also reflected in the magnetic field dependence @xcite . if one studies @xmath62 for a given interaction strength but increasing magnetic fields , the system has a non - monotonic evolution .
examples of this behavior are shown in ref.@xcite .
for low magnetic field ( @xmath78 t , and ` unit ' interaction , with @xmath79 ) , @xmath62 is not the harmonic oscillator delta - like function , but is somewhat shifted towards a gue ( brody - fit with @xmath80 ) .
as the field increases , however , the @xmath18-distribution reaches a ` maximum ' crossover ( @xmath81 , for @xmath82 t ) , before going back towards a more delta - like function , as seen for @xmath83 t , which drops faster than a poissonian distribution . that this occurs is understandable , since for high magnetic fields one would expect to reach a regime where the interactions would be a weak perturbation ( for a given value of @xmath50 ) , and the spectrum would evolve towards a set of landau levels
this non - monotonic behavior is seen in all the level sequences we have studied .
we have shown that the interactions introduce classical non - integrability in a quantum - dot system , even if the geometry is integrable for the one - particle problem . moreover , this behavior , given the bohigas _
et al_. conjecture @xcite , would be expected to be reflected in the accompanying level structure of the corresponding quantum mechanical version of the system .
we have found , that as we introduce interactions , indeed the nns fluctuation distribution function exhibits an apparent crossover towards one of the rmt functions , with a characteristic maximum away from zero spacing .
this crossover , is found to be not complete , however , even for rather strong interaction , perhaps due to the mixed dynamics in the classical counterpart .
the nns distribution exhibits non - monotonic magnetic field dependence , associated with the fact that in that regime the particle interactions are but a weak perturbation of the landau level spectrum .
a number of theoretical questions still remain , however , as more direct comparison of the classical and quantum systems is needed , perhaps including a study of the associated wave functions and possible appearance of ` scars ' @xcite .
nevertheless , the importance of interactions in the details of level structure and associated experiments , as discussed in the introduction , can not be ignored .
we hope that this , as well as semiclassical treatments ( see richter s article in these proceedings ) , motivate further understanding of this fascinating problem .
gorkov and g.m .
eliashberg , sov .
jetp * 21 * , 940 ( 1965 ) ; k.b .
efetov , adv . phys . * 32 * , 53 ( 1983 ) ; b.l .
altshuler and b.z .
spivak , sov .
jetp * 65 * , 343 ( 1987 ) ; y.h .
zheng and r.a .
serota , phys .
b * 50 * , 2492 ( 1994 ) . | the role of electronic interactions in the level structure of semiconductor quantum dots is analyzed in terms of the correspondence to the integrability of a classical system that models these structures .
we find that an otherwise simple system is made strongly non - integrable in the classical regime by the introduction of particle interactions .
in particular we present a two - particle classical system contained in a @xmath0-dimensional billiard with hard walls .
similarly , a corresponding two - dimensional quantum dot problem with three particles is shown to have interesting spectral properties as function of the interaction strength and applied magnetic fields .
+ keywords : quantum dots , chaos , level statistics + sergio e. ulloa + e - mail : [email protected] + fax : ( 614 ) 5930433 |
shortly after the big bang , large density fluctuations in the early universe may have resulted in formation of primordial black holes ( pbhs ) @xcite .
there is a wide range of allowed masses for pbhs .
depending on the epoch and conditions during formation , pbh masses can be anywhere from approximately a gram to a million solar masses .
while low mass pbhs would have already evaporated through hawking radiation @xcite , large ones with masses @xmath4 g would still be present today .
it is also possible to have prolonged pbh formation during a non - radiation - dominated phase of the universe where pbhs can form with a continuum mass distribution , rather than mostly at one particular mass scale as in the conventional radiation dominated case @xcite .
stable pbhs can be cosmologically - significant , and may serve as an ideal dark matter candidate @xcite . depending on the pbh abundance , hawking radiation from pbhs with lifetime longer than the age of the universe may be observable .
extragalactic gamma rays strongly constrain pbhs in the mass range @xmath5 g @xcite .
pbhs in the mass range @xmath6 g are bounded by femtolensing of gamma - ray bursts @xcite , and pbhs in the mass range @xmath7 m@xmath8 are constrained by gravitational microlensing @xcite .
in addition , pbhs with mass @xmath9 m@xmath8 may be constrained by the accretion of matter in the early universe @xcite .
black holes at approximately this mass are constrained by x - rays observations @xcite .
for recent comprehensive reviews on astrophysical constraints on pbhs see refs .
@xcite .
in this paper we present a new bound on pbhs in the mass range @xmath5 g using the most recent planck cosmic microwave background ( cmb ) data @xcite .
the cmb is sensitive to additional sources of energy injection during the recombination epoch , which leads to damping of the anisotropies . for pbhs ,
this energy injection is due to hawking radiation . as we show
, the planck data now place a stronger bound on pbhs over a larger fraction of this mass range than previous most stringent bounds derived from the extragalactic gamma - ray background ( egb ) .
previous authors have used planck data to bound pbhs in the mass regime that we study @xcite ; as discussed below we precisely identify the mass regime over which the cmb and egb bounds are dominant .
in addition we note that our analysis is distinct from previous studies that used early - time distortions of the cmb to bound pbhs in the mass range @xmath10 g @xcite .
the theoretical formalism that we utilize to constrain pbhs is similar to that used to constrain dark matter annihilation or decay @xcite . from the perspective of the cmb ,
pbh evaporation is most similar to dark matter decay in that the energy injection only depends on the pbh mass and abundance , and is at a steady rate to the present time .
this energy injection can have a significant impact on ionization at low redshift .
however unlike the energy from dark matter , which can be injected in the form of heavy standard model particles , pbhs with mass @xmath11 g mostly radiate in electrons and photons , or other ( near ) massless species , but generally not into much heavier particle species .
because pbhs with mass @xmath12 g are too cold to emit electrons , their injection into the cmb is unobservable , and the bounds are weak above this mass . this paper is organized as follows . in section [ blackholes ]
we briefly discuss the injection of radiation from pbhs . in section [ igm_interactions ]
we discuss the impact of this injected energy on the igm . in section [ recomb_hist ] ,
we show the modifications on the ionization history , and in section [ cmb_alter ] we discuss cmb distortions . in section [ black_hole_constr ]
we show the resulting constraints on pbhs , and section [ conclusions ] presents the discussion and conclusions .
depending on their mass , pbhs radiate a spectrum of particles , which decay via cascades into photons , electrons , protons , and neutrinos . these particles then deposit energy into the igm .
the injection of energy is described by the equation @xmath13 where @xmath14 is the pbh number density , @xmath15 is the critical density of the universe today , @xmath16 is the pbh mass , @xmath17 is the pbh density observed today relative to the critical density .
note that the above equations assume that pbhs are comprised of a single mass and the mass does not changes as it radiates .
this is satisfied as long as the lifetime is large compared to the age of the universe and is satisfied by the masses considered in this work .
apart from some cosmetic differences , equation ( [ equ : bh - energy ] ) is identical to that for decaying dark matter @xcite . in order to evaluate equation ( [ equ : bh - energy ] ) , an expression for @xmath18 is required . to obtain this , we start from the fact that hawking radiation equates the radiation from a black hole to the blackbody radiation of an object with temperature @xmath19 and with an emission spectrum @xmath20 where @xmath21 is the spin of the radiated particle and @xmath22 is the absorption coefficient for the particle . for low @xmath23
the absorption coefficient can deviate greatly from the geometric optic limit @xcite , @xmath24 for standard model particles , the average radiated energy in the massless limit is @xcite @xmath25 for pbhs with masses in the range @xmath26 grams , the fraction of emitted particles of different spins is @xcite @xmath27 it should be noted that these fractions are not normalized to unity , but rather to a @xmath28 g pbh . with this information
the energy injection in equation ( [ equ : bh - energy ] ) can be evaluated after using @xcite @xmath29 since only the electrons and photons interact electromagnetically for the pbh masses considered , these are the only fractions that need to be considered for calculating the energy output .
the energy deposited into the igm by pbhs is absorbed through multiple channels .
following previous studies , here we consider three channels for the igm interaction : hydrogen ionization , lyman - alpha excitations , and heating the igm @xcite .
these effects alter the cosmological recombination equations as @xmath30 where @xmath31 is the helium fraction , @xmath32 is the ionization fraction , @xmath33 is the boltzmann constant , and @xmath34 is the hubble parameter .
the standard equations without additional energy injection from pbhs are denoted by the subscript orig " and are derived in e.g. ref .
the quantities @xmath35 , @xmath36 , and @xmath37 are factors corresponding to the additional energy injection affecting ionization from the ground state , ionization from excited states , and heating the igm .
each of these injections are dependent on the injection energy through @xmath38 here @xmath39 is the hydrogen number density , and @xmath40 and @xmath41 are the energies of the ground and the excited hydrogen atom electron levels respectfully . the quantity @xmath42 is related to the probability for an excited hydrogen atom to emit a photon prior to being ionized @xcite .
the quantities @xmath43 , @xmath44 , @xmath45 are efficiencies for energy interactions through each channel . commonly referred to as effective efficiencies , they are redshift , energy , and species dependent quantities that equate the total energy injection to the actual amount absorbed through a pathway @xcite .
previously , these efficiencies have been approximated by simple @xmath32 dependent equations , with the energy injection taken to be instantaneous , through a technique known as the `` ssck '' method , described in further detail in ref .
@xcite . to calculate the effective efficiencies , we follow the approach adopted in ref .
these efficiencies have been tabulated for electron and photon particle injection into the igm at various redshifts and particle energies , and divided into five different channels : hydrogen ionization , helium ionization , lyman - alpha excitations , heating , and continuum photons ( energy lost as photons with @xmath46 ) @xcite .
the hydrogen , lyman - alpha , and heating efficiencies were used to calculate the various efficiencies in equations ( [ equ : i_xi ] ) , ( [ equ : i_xalpha ] ) , and ( [ equ : k_h ] ) .
we do not consider helium ionization because it is subdominant .
we additionally note that we do not consider energy deposited into continuum photons .
this is because continuum photon energy affects the cmb mostly through spectral distortions rather than through anisotropies , and the anisotropies are the focus of this paper .
the impact on spectral distortions is also less significant than on anisotropies @xcite ; a basic demonstration of this is given in the appendix
. refs .
@xcite tabulate efficiencies specifically for decaying and annihilating dark matter into electron and photon channels .
these efficiencies are specifically valid for a small ionization fraction , in which case the efficiency scales linearly with ionization fraction . for dark matter decay ,
the energy injection is linearly proportional to the density , and for annihilation it is proportional to the density - squared . from equation [ equ : bh - energy ] , the radiation from pbhs is linearly proportional to density , therefore this is most analogous to dark matter decay .
for this reason , we utilize the decay efficiencies to produce black hole efficiencies . to obtain the pbh efficiency
, we assume that pbh injected particles have the same efficiency as those with the average energy for its species .
combining the different species efficiencies through a weighted average based upon their emission fractions gives @xmath47 using the decay efficiencies from refs .
@xcite , figure [ fig : effective_efficiency ] shows the effective efficiencies for pbh hawking radiation for the mass range considered in this work .
it should be noted that there appears a location where the efficiency drops drastically for all channels used in altering the ionization history , and this location shifts to later times for increasing temperatures .
this drop will have a direct impact on the constraints for a given pbh mass . [ cols="^,^ " , ] since there is little variation in the base cosmological parameters , for computational convenience to set upper limits on @xmath17 we take the six principle cosmological parameters to be fixed at their best fit values in the case of no additional energy injection @xcite .
figure [ fig : egb_comparison ] shows the result of the 95% confidence limit , where the confidence limit is defined as the cumulative distribution centered around the median , which corresponds closely to the peak of the distributions in figure [ fig : egb_comparison ] .
the constraint follows the expected inverse cube relationship to the pbh mass which is predicted by the energy injection formula . at the 95% confidence level ( shaded region )
compared with the same exclusion bound enforced by egb , assuming 100% of the background produced by pbhs ( long dashes ) .
also included is an estimation for the bound that is imposed due to femtolensing ( short dashes ) considered in @xcite . ] in addition to the cubic dependence on mass , there is also a highly nonlinear relationship to the effective efficiency .
this nonlinearity is most prevalent at a pbh mass around @xmath48 g. comparing effective efficiencies , in figure [ fig : effective_efficiency ] , the trend is correlated with the efficiency values that occur near the time of recombination . as the efficiency value decreases , it is required for a larger amount of total energy to be created in order to produce the same effect .
for this reason , as the efficiency experiences a large decrease , the allowable maximum mass fraction increases . we note that since the six base cosmological parameters were fixed , the pbh abundance may be more strongly constrained than in a model in which more parameters are allowed to vary . to check this , we compare to the case in which the base cosmological parameters are allowed to vary for a single pbh mass of @xmath49 g. we find that by freeing all of the cosmological parameters , the constraint on @xmath17 may be weakened by up to a factor of three .
however , as stated above for computational convenience we have decided to fix the base cosmological parameters for our main bounds on @xmath50 . as indicated above previous analyses
have used the egb to constrain pbhs in the mass regime @xmath0 g @xcite .
since pbhs emit a mass - dependent gamma - ray spectrum , there is an upper bound on their density before they would be excluded by egb measurements . following the prescription outlined in ref .
@xcite , the number density of photons @xmath51 with energy @xmath52 and their intensity is @xmath53 @xmath54 where @xmath55 is the time when photon creation begins .
the quantity @xmath56 is the photon spectrum given by equation ( [ equ : hawking - dist ] ) , which we take at the high energy limit .
for pbhs in the mass range studied , peak intensity occurs at @xmath57 mev .
constraints were derived by matching the intensity to the upper bound of the comptel egb experimental data @xcite .
egb constraints are also shown in figure [ fig : egb_comparison ] as well as those imposed by femtolensing @xcite .
we find that planck provides the strongest constraint on the abundance of pbhs for masses @xmath1 g , while the egb dominates for masses @xmath2 g. note that this conclusion differs from that of ref .
the planck constraint deviates from a linear relation because of the model for effective efficiencies .
pbhs are of great interest in cosmology .
they reveal conditions in the early universe and can serve as a dark matter candidate .
there are several standard mechanisms that have been proposed to detect pbhs ; these include detection of hawking radiation , detection of radiation produced from accretion disks , and gravitational lensing .
each method is capable of targeting different pbh mass ranges . in this paper , we have focused on pbhs with masses in the range @xmath0 g. we have improved and made more precise the constraints in this mass range using the cmb and egb . for our cmb bound , we model the energy absorption not as instantaneous , but rather using redshift dependent efficiency . the energy injection results in an increase in the ionization fraction at late times as well as an increase in the igm temperature , leading to distortions of the cmb anisotropies .
larger fractional changes occur at large multipoles because of the increase of the width of the last scattering surface . using planck data ,
we show that cmb distortions from hawking radiation allow for stringent constraints on the density of @xmath0 g pbhs of @xmath58 .
we show that for mass @xmath1 g , the cmb constraints are stronger than the constraints from the @xmath59 mev egb , which imply , @xmath60 .
constraints imposed by cmb spectral distortions from hawking radiation producing sub 10.2 ev photons are also much weaker than our constraint . in the future
, our theoretical analysis may be improved by including a mass spectrum of pbhs .
in addition , even though we have used the egb to bound the contribution of pbhs , it may be interesting to consider the egb as a signal of pbhs .
this is an exciting possibility because the origin of this @xmath61 mev gamma - ray background is not yet known @xcite .
future missions to measure mev gamma - rays will be especially important for the study of pbhs @xcite .
the authors thank daniel meerburg for helpful discussion and modified hyrec codes .
we also thank tomohiro harada and tracy slatyer for helpful discussions .
bd acknowledges support from doe grant de - fg02 - 13er42020 .
les acknowledges support from nsf grant phy-1522717 .
yg thanks the mitchell institute for fundamental physics and astronomy and wayne state university for support .
sc acknowledges support from nasa astrophysics theory grant nnx12ac71 g .
sw is supported in part by nasa astrophysics theory grant nnh12zda001n and doe grant de - fg02 - 85er40237 .
99 s. hawking , mon . not .
soc . * 152 * , 75 ( 1971 ) .
y. b. zeldovich , i. d. novikov , astron .
volume 43 , pages 758 , 1966 ; y. b. zeldovich , i. d. novikov , sov .
volume 10 , pages 602 , 1967 .
b. j. carr and s. w. hawking , mon . not .
* 168 * , 399 ( 1974 ) .
s. w. hawking , nature * 248 * , 30 ( 1974 ) .
doi:10.1038/248030a0 j. georg , g. sengor , and s. watson , phys . rev .
d * 93 * , no .
12 , 123523 ( 2016 ) doi:10.1103/physrevd.93.123523 [ arxiv:1603.00023 [ hep - ph ] ] . t. harada , c. m. yoo , k. kohri , k. i. nakao and s. jhingan , astrophys . j. * 833 * , no . 1 , 61 ( 2016 ) doi:10.3847/1538 - 4357/833/1/61 [ arxiv:1609.01588 [ astro-ph.co ] ] . b. j. carr , k. kohri , y. sendouda and j. yokoyama , phys . rev .
d * 81 * , 104019 ( 2010 ) doi:10.1103/physrevd.81.104019 [ arxiv:0912.5297 [ astro-ph.co ] ] .
r. j. nemiroff , g. f. marani , j. p. norris and j. t. bonnell , phys .
lett . * 86 * , 580 ( 2001 ) doi:10.1103/physrevlett.86.580 [ astro - ph/0101488 ] .
k. griest , a. m. cieplak and m. j. lehner , phys .
lett . * 111 * , no .
18 , 181302 ( 2013 ) .
doi:10.1103/physrevlett.111.181302 m. ricotti , j. p. ostriker and k. j. mack , astrophys . j. * 680 * , 829 ( 2008 ) doi:10.1086/587831 [ arxiv:0709.0524 [ astro - ph ] ]
. l. chen , q. g. huang and k. wang , arxiv:1608.02174 [ astro-ph.co ] .
y. ali - hamoud and m. kamionkowski , arxiv:1612.05644 [ astro-ph.co ] .
d. gaggero , g. bertone , f. calore , r. m. t. connors , m. lovell , s. markoff and e. storm , arxiv:1612.00457 [ astro-ph.he ] . b. carr , f. kuhnel and m. sandstad , phys . rev .
d * 94 * , no .
8 , 083504 ( 2016 ) doi:10.1103/physrevd.94.083504 [ arxiv:1607.06077 [ astro-ph.co ] ] .
a. m. green , phys .
d * 94 * , no .
6 , 063530 ( 2016 ) doi:10.1103/physrevd.94.063530 [ arxiv:1609.01143 [ astro-ph.co ] ]
. n. aghanim _ et al . _
[ planck collaboration ] , astron .
astrophys . * 594 * , a11 ( 2016 ) doi:10.1051/0004 - 6361/201526926 [ arxiv:1507.02704 [ astro-ph.co ] ] .
v. poulin , j. lesgourgues and p. d. serpico , arxiv:1610.10051 [ astro-ph.co ] .
h. tashiro and n. sugiyama , phys .
d * 78 * , 023004 ( 2008 ) doi:10.1103/physrevd.78.023004 [ arxiv:0801.3172 [ astro - ph ] ] .
l. zhang , x. chen , m. kamionkowski , z. g. si and z. zheng , phys .
d * 76 * , 061301 ( 2007 ) doi:10.1103/physrevd.76.061301 [ arxiv:0704.2444 [ astro - ph ] ] .
m. s. madhavacheril , n. sehgal and t. r. slatyer , phys .
d * 89 * , 103508 ( 2014 ) doi:10.1103/physrevd.89.103508 [ arxiv:1310.3815 [ astro-ph.co ] ] . t. r. slatyer , phys .
d * 93 * , no .
2 , 023527 ( 2016 ) doi:10.1103/physrevd.93.023527 [ arxiv:1506.03811 [ hep - ph ] ] . t. r. slatyer , phys .
d * 93 * , no . 2 , 023521 ( 2016 ) doi:10.1103/physrevd.93.023521 [ arxiv:1506.03812 [ astro-ph.co ] ]
. h. liu , t. r. slatyer and j. zavala , phys .
d * 94 * , no . 6 , 063507 ( 2016 ) doi:10.1103/physrevd.94.063507 [ arxiv:1604.02457 [ astro-ph.co ] ] . t. r. slatyer and c. l. wu , arxiv:1610.06933 [ astro-ph.co ] . j. h. macgibbon and b. r. webber , phys . rev .
d * 41 * , 3052 ( 1990 ) .
doi:10.1103/physrevd.41.3052 ; j. h. macgibbon , phys .
d * 44 * , 376 ( 1991 ) .
doi:10.1103/physrevd.44.376 j. s. bolton , g. d. becker , s. raskutti , j. s. b. wyithe , m. g. haehnelt and w. l. w. sargent , doi:10.1111/j.1365 - 2966.2011.19929.x arxiv:1110.0539 [ astro-ph.co ] . j. s. bolton , g. d. becker , j. s. b. wyithe , m. g. haehnelt and w. l. w. sargent , mon . not .
406 * , 612 ( 2010 ) doi:10.1111/j.1365 - 2966.2010.16701.x [ arxiv:1001.3415 [ astro-ph.co ] ] .
k. m. belotsky and a. a. kirillov , jcap * 1501 * , no . 01 , 041 ( 2015 ) doi:10.1088/1475 - 7516/2015/01/041 [ arxiv:1409.8601 [ astro-ph.co ] ] . t.
r. slatyer , phys .
d * 87 * , no .
12 , 123513 ( 2013 ) doi:10.1103/physrevd.87.123513 [ arxiv:1211.0283 [ astro-ph.co ] ] . y. ali - haimoud and c. m. hirata , phys .
d * 83 * , 043513 ( 2011 ) doi:10.1103/physrevd.83.043513 [ arxiv:1011.3758 [ astro-ph.co ] ] .
a. lewis , a. challinor and a. lasenby , astrophys .
j. * 538 * , 473 ( 2000 ) doi:10.1086/309179 [ astro - ph/9911177 ] . c. howlett , a. lewis , a. hall and a. challinor , jcap * 1204 * , 027 ( 2012 ) doi:10.1088/1475 - 7516/2012/04/027 [ arxiv:1201.3654 [ astro-ph.co ] ] .
a. lewis , phys .
d * 87 * , no .
10 , 103529 ( 2013 ) doi:10.1103/physrevd.87.103529 [ arxiv:1304.4473 [ astro-ph.co ] ] .
a. lewis and s. bridle , phys .
d * 66 * , 103511 ( 2002 ) doi:10.1103/physrevd.66.103511 [ astro - ph/0205436 ] .
n. padmanabhan and d. p. finkbeiner , phys .
d * 72 * , 023508 ( 2005 ) doi:10.1103/physrevd.72.023508 [ astro - ph/0503486 ] .
`` planck 2015 results : cosmological parameter tables , '' g. weidenspointner .
the origin of the cosmic gamma - ray background in the comptel energy range .
phd thesis , technical university of munich , munich , germany ?
id=602832 .
j. zavala , m. vogelsberger and s. d. m. white , phys .
d * 81 * , 083502 ( 2010 ) doi:10.1103/physrevd.81.083502 [ arxiv:0910.5221 [ astro-ph.co ] ] .
l. e. strigari , j. f. beacom , t. p. walker and p. zhang , jcap * 0504 * , 017 ( 2005 ) doi:10.1088/1475 - 7516/2005/04/017 [ astro - ph/0502150 ] .
p. ruiz - lapuente , l. s. the , d. hartmann , m. ajello , r. canal , f. k. ropke , s. t. ohlmann and w. hillebrandt , astrophys .
j. * 820 * , no . 2 , 142 ( 2016 ) doi:10.3847/0004 - 637x/820/2/142 [ arxiv:1502.06116 [ astro-ph.he ] ] . s. horiuchi and j. f. beacom , astrophys .
j. * 723 * , 329 ( 2010 ) doi:10.1088/0004 - 637x/723/1/329 [ arxiv:1006.5751 [ astro-ph.co ] ] .
a. de angelis _ et al . _
[ e - astrogam collaboration ] , arxiv:1611.02232 [ astro-ph.he ] .
continuum photons affect the cmb by creating spectral distortions @xcite . ref .
@xcite investigated limits to these spectral distortions modeling the distortions as a bose - einstein distribution with a chemical potentioal @xmath62 , @xmath62-type distortions . in order to get a baseline estimate on the effect of the continuum photons injected by pbh on the cmb
a similar approach was taken . assuming the injection will alter the perfect blackbody spectrum with the same @xmath62-type distortions
, these distortions will be approximately where @xmath64 and @xmath65 are the energy density of the cmb and the distortion injection rate on the cmb respectfully .
@xmath65 becomes the injection rate directly from the black hole energy going into the continuum .
the result is current limits on these distortions are @xmath68 at two sigma @xcite .
a comparison of this constraint and the cmb value discussed in this work is shown in figure [ fig : spectral_distortion_comparison ] . as can be seen ,
the constraints produced by this work are much stronger , several orders of magnitude , than limits produced assuming a @xmath62-type or similar distortion .
the only limits that approache the cmb result are those that consider alterations at extremely late times .
these however can be ignored due to cmb photons at this time being much cooler than the 10.2 ev limit for these injections .
thus , the actual alterations at this time is further reduced .
confidence level compared with the same exclusion bound produced through spectral distortions .
the various curves correspond to different integration limits in equation ( [ equ : spec_dist_mu ] ) .
the most constraining uses @xmath69 to infinity , and the second uses @xmath70 to infinity . ] | we investigate constraints on the abundance of primordial black holes ( pbhs ) in the mass range @xmath0 g using data from the cosmic microwave background ( cmb ) and mev extragalactic gamma - ray background ( egb ) .
hawking radiation from pbhs with lifetime greater than the age of the universe leaves an imprint on the cmb through modification of the ionization history and the damping of cmb anisotropies . using a model for redshift dependent energy injection efficiencies ,
we show that a combination of temperature and polarization data from planck provides the strongest constraint on the abundance of pbhs for masses @xmath1 g , while the egb dominates for masses @xmath2 g. both the cmb and egb now rule out pbhs as the dominant component of dark matter for masses @xmath3 g. planned mev gamma - ray observatories are ideal for further improving constraints on pbhs in this mass range .
mi - th-1636 , wsu - hep-1609 |
today s state - of - the - art perceptual models @xcite have mostly tackled detecting and recognizing individual objects _ in isolation_. however , understanding a visual scene often goes beyond recognizing individual objects .
take a look at the two images in fig .
[ fig : intro_figure ] .
even a perfect object detector would struggle to perceive the subtle difference between a man feeding a horse and a man standing by a horse .
the rich semantic relationships between these objects have been largely untapped by these models . as indicated by a series of previous works @xcite , one crucial step towards a deeper understanding of visual scenes is building a structured representation that captures objects and their semantic relationships .
such representation not only offers contextual cues for fundamental recognition tasks @xcite but also provide values in a larger variety of high - level visual tasks @xcite . the recent success of deep learning - based recognition models @xcite has resurged interest in examining the detailed structures of a visual scene , especially in the form of object relationships @xcite . scene graph , proposed by johnson @xcite , offers a platform to explicitly model objects and their relationships . in short , a _ scene graph _ is a visually - grounded graph over the object instances in an image , where the edges depict their pairwise relationships ( see example in fig .
[ fig : intro_figure ] ) .
the value of scene graph representation has been proven in a wide range of visual tasks , such as semantic image retrieval @xcite , 3d scene synthesis @xcite , and visual question answering @xcite .
anderson recently proposed spice @xcite as an enhanced automated caption evaluation metric defined over scene graphs .
however , these models that use scene graphs either rely on ground - truth annotations @xcite , synthetic images @xcite , or extract a scene graph from text domain @xcite . to truly take advantage of such rich structure , it is crucial to devise a model that automatically generates scene graphs from images .
[ fig : intro_figure ] in this work , we address the problem of scene graph generation , where the goal is to generate a visually - grounded scene graph from an image . in a generated scene graph ,
an object instance is characterized by a bounding box with an object category label , and a relationship is characterized by a directed edge between two bounding boxes ( i.e. , object and subject ) with a relationship predicate ( red nodes in fig .
[ fig : intro_figure ] ) . the major challenge of generating scene graphs is reasoning about relationships .
much effort has been expended on localizing and recognizing semantic relationships in images @xcite .
most methods have focused on making _
local _ predictions of object relationships @xcite , which essentially simplify the scene graph generation problem into independently predicting relationships between pairs of objects . however , by doing _ local
_ predictions these models ignore surrounding context , whereas joint reasoning with contextual information can often resolve ambiguity due to local predictions in isolation . to capture this intuition
, we propose a novel end - to - end model that learns to generate image - grounded scene graphs ( fig .
[ fig : pipeline ] ) .
the model takes an image as input and outputs a scene graph that consists of object categories , their bounding boxes , and semantic relationships between pairs of objects .
our major contribution is that instead of inferring each component of a scene graph in isolation , the model passes messages containing contextual information between a pair of bipartite sub - graphs of the scene graph , and iteratively refines its predictions using rnns .
we evaluate our model on the visual genome scene graph dataset @xcite , which contains human - annotated scene graphs on 108,077 images . on average , each image
is annotated with 13.5 objects and 15 pairwise object relationships .
we show that relationship prediction in scene graphs can be significantly improved by our model .
furthermore , we also apply our model to the nyu depth v2 dataset @xcite , establishing new state - of - the - art results in reasoning about spatial relations , such as horizontal and vertical supports . in summary , we propose an end - to - end model that generates visually - grounded scene graphs from images .
the model uses a novel inference formulation that iteratively refines its prediction by passing contextual messages along the topological structure of a scene graph .
we demonstrate its use for generating semantic scene graphs from the visual genome scene graph dataset as well as predicting support relations using the nyu depth v2 dataset @xcite .
* scene understanding and relationship prediction .
* visual scene understanding often harnesses the statistical patterns of object co - occurrence @xcite as well as spatial layout @xcite .
a series of contextual models based on surrounding pixels and regions have also been developed for perceptual tasks @xcite .
recent works @xcite exploits more complex structures for relationship prediction .
however , these works focus on image - level predictions without detailed visual grounding . physical relationships , such as support and stability , have been studied in @xcite .
lu @xcite directly tackled the semantic relationship detection by combining visual inputs with language priors to cope with the long - tail distribution of real - world relationships .
however , their method predicts each relationship independently .
we show that our model outperforms theirs with joint inference
. * visual scene representation .
* one of the most popular ways of representing a visual scene is through text descriptions @xcite .
although text - based representation has been shown to be helpful for scene classification and retrieval , its power is often limited by ambiguity and lack of expressiveness . in comparison ,
scene graphs @xcite offer explicit grounding of visual concepts , avoiding referential uncertainty in text - based representation .
scene graphs have been used in many downstream tasks such as image retrieval @xcite , 3d scene synthesis @xcite and understanding @xcite , visual question answering @xcite , and automatic caption evaluation @xcite . however , previous work on scene graphs shied away from the graph generation problem by either using ground - truth annotations @xcite , or extracting the graphs from other modalities @xcite .
our work addresses the problem of generating scene graphs directly from images
. * graph inference . * conditional random fields ( crf ) have been used extensively in graph inference .
johnson used crf to infer scene graph grounding distributions for image retrieval @xcite .
yatskar @xcite proposed situation - driven object and action prediction using a deep crf model .
our work is closely related to crfasrnn @xcite and graph - lstm @xcite in that we also formulate the graph inference problem using an rnn - based model .
a key difference is that they focus on node inference while treating edges as pairwise constraints , whereas we enable edge predictions using a novel primal - dual graph inference scheme .
we also share the same spirit as structural rnn @xcite .
a crucial distinction is that our model iteratively refines its predictions through message passing , whereas the structural rnn model only makes one - time predictions along the temporal dimension , and thus can not refine its past predictions .
a _ scene graph _ , as defined by johnson @xcite , is a structured representation of an image , where nodes in a scene graph correspond to object bounding boxes with their object categories , and edges correspond to their pairwise relationships between objects .
the task of _ scene graph generation _ is to generate a visually - grounded scene graph that most accurately correlates with an image .
intuitively , individual predictions of objects and relationships can benefit from their surrounding context .
for instance , knowing `` a horse is on grass field '' is likely to increase the chance of detecting a person and predicting the relationship of `` man riding horse '' . to capture this intuition , we propose a joint inference framework to enable contextual information to propagate through the scene graph topology via a message passing scheme .
however , inference on a densely connected graph can be very expensive .
as shown in previous work @xcite and @xcite , dense graph inference can be approximated by mean field in conditional random fields ( crf ) .
our approach is inspired by zeng @xcite , which designs fully differentiable layers to enable end - to - end learning with recurrent neural networks ( rnn ) .
yet their model relies on purpose - built rnn layers . to achieve greater flexibility in a more principled training framework
, we use a generic rnn unit instead , in particular a gated recurrent unit ( gru ) @xcite . at each iteration
, each gru takes its previous hidden state and an incoming message as input , and produces a new hidden state as output .
each node and edge in the scene graph maintains its internal state in its corresponding gru unit , where all nodes share the same gru weights ( node grus ) , and all edges share the other set of gru weights ( edge grus ) .
this setup allows the model to pass messages ( an aggregation of gru hidden states ) among the gru units along the scene graph topology .
we also propose a message pooling function that learns to dynamically aggregate the hidden states of the grus into messages .
we further observe that the unique structure of scene graphs forms a bipartite structure of message passing channels . since messages only pass along the topological structure of a scene graph , the set of edge grus and the set of node grus form a bipartite graph , where no message is passed inside each set .
inspired by this observation , we formulate two disjoint sub - graphs that are essentially the dual graph to each other .
the primal graph defines channels for messages to pass from edge grus to node grus .
the dual graph defines channels for messages to pass from node grus to edge grus . with such primal - dual formulation
, we can therefore improve inference efficiency by iteratively passing messages between these sub - graphs instead of through a densely connected graph .
[ fig : architecture ] gives an overview of our model .
we first lay out the mathematical formulation of our scene graph generation problem . to generate a visually grounded scene graph
, we need to obtain an initial set of object bounding boxes .
these bounding boxes can be either from ground - truth human annotation or algorithmically generated . in practice , we use the region proposal network ( rpn ) @xcite to automatically generate a set of object bounding box proposals @xmath0 from an image @xmath1 as the base input to the inference procedure ( fig .
[ fig : architecture](a ) ) . for each object box proposal
, we need to infer two types of object - centric variables : 1 ) an object class label , and 2 ) four bounding box offsets relative to the proposal box coordinates , which are used for refining the proposal boxes .
in addition , we need to infer a relationship - centric variable between every pair of proposal boxes , which denotes the predicate type of the relationship between the corresponding object pair . given a set of object classes @xmath2 ( including background ) and a set of relationship types @xmath3 ( including none relationship ) , we denote the set of all variables to be @xmath4 , where @xmath5 is the number of proposal boxes , @xmath6 is the class label of the @xmath7-th proposal box , @xmath8 is the bounding box offsets relative to the @xmath7-th proposal box coordinates , and @xmath9 is the relationship predicate between the @xmath7-th and the @xmath10-th proposal boxes . at the high level ,
the inference task is to classify objects , predict their bounding box offsets , and classify relationship predicates between each pair of objects .
formally , we formulate the scene graph generation problem as finding the optimal @xmath11 that maximizes the following probability function given the image @xmath1 and box proposals @xmath0 : @xmath12 in the next subsection , we introduce a way to approximate the inference procedure using an iterative message passing scheme modeled with gated recurrent units @xcite .
we use mean field to perform approximate inference .
we denote the probability of each variable @xmath13 as @xmath14 , and assume that the probability only depends on the current state of each node and edge at each iteration .
in contrast to zeng @xcite , we use a generic rnn module to compute the hidden states .
in particular , we choose gated recurrent units @xcite due to its simplicity and effectiveness .
we use the hidden state of the corresponding gru , a high - dimensional vector , to represent the current state of each node and each edge . as all the nodes ( edges )
share the same update rule , we share the same set of parameters among all the node grus , and the other set of parameters among all the edge grus ( fig . [
fig : architecture ] ) .
we denote the current hidden state of node @xmath7 as @xmath15 and the current hidden state of edge @xmath16 as @xmath17 .
then the mean field distribution can be formulated as @xmath18 where @xmath19 is the visual feature of the @xmath7-th node , and @xmath20 is the visual feature of the edge from the @xmath7-th node to the @xmath10-th node . in the first iteration ,
the gru units take the visual features @xmath21 and @xmath22 as input ( fig .
[ fig : architecture](a ) ) .
we use the visual feature of the proposal box as the visual feature @xmath19 for the @xmath7-th node .
we use the visual feature of the union box over the proposal boxes @xmath23 as the visual feature @xmath24 for edge @xmath25 .
these visual features are extracted by a roi - pooling layer @xcite from the image . in later iterations ,
the inputs are the aggregated messages from other gru units of the previous step .
we talk about how the messages are aggregated and passed in the next subsection .
[ sec : rnn ] offers a generic formulation for solving graph inference problem using rnns .
however , we observe that we can further improve the inference efficiency by leveraging the unique bipartite structure of a scene graph . in the scene graph topology ,
the neighbors of the edge grus are node grus , and vice versa . passing messages along this
structure forms two disjoint sub - graphs that are the dual graph to each other .
specifically , we have a node - centric primal graph , in which each node gru gets messages from its inbound and outbound edge grus . in the edge - centric dual graph
, each edge gru gets messages from its subject node gru and object node gru ( fig . [
fig : architecture](b ) ) .
we can therefore improve inference efficiency by iteratively passing messages between these two sub - graphs instead of through a densely connected graph ( fig .
[ fig : architecture](c ) ) .
as each gru receives multiple incoming messages , we need an aggregation function that can fuse information from all messages into a meaningful representation
. a nave approach would be standard pooling methods such as average- or max - pooling .
however , we found that it is more effective to learn adaptive weights that can modulate the influences of incoming messages and only keep the relevant information .
we introduce a _ message pooling _ function that computes the weight factors for each incoming message and fuse the messages using a weighted sum .
we provide an empirical analysis of different message pooling functions in sec .
[ sec : experiment ] .
formally , given the current gru hidden states of nodes and edges @xmath15 and @xmath17 , we denote the messages to update the @xmath7-th node as @xmath26 , which is computed by a function of its own hidden state @xmath15 , and the hidden states of its outbound edge grus @xmath27 and inbound edge grus @xmath28 .
similarly , we denote the message to update the edge from the @xmath7-th node to the @xmath10-th node as @xmath29 , which is computed by a function of its own hidden state @xmath27 , the hidden states of its subject node gru @xmath15 and its object node gru @xmath30 . to be more specific , @xmath26 and @xmath31
are computed by the following two adaptively weighted message pooling functions : @xmath32 ) h_{i\rightarrow j } + \sum_{j : j \rightarrow i } \sigma(\mathbf{v}_2^t[h_i , h_{j\rightarrow i } ] ) h_{j\rightarrow i } \label{eq : vert}\\ & m_{i\rightarrow j } = \sigma(\mathbf{w}_1^t[h_i , h_{i\rightarrow j } ] ) h_{i } + \sigma(\mathbf{w}_2^t[h_j , h_{i\rightarrow j } ] ) h_{j } \label{eq : edge}\end{aligned}\ ] ] where @xmath33 $ ] denotes a concatenation of vectors , and @xmath34 denotes a sigmoid function . @xmath35 , @xmath36 and @xmath37 , @xmath38 are learnable parameters .
these two equations describe the primal - dual update rules , as shown in ( b ) of fig .
[ fig : architecture ] .
our final output layers follow closely with the faster r - cnn setup @xcite .
we use a softmax layer to produce the final scores for the object class as well as relationship predicate .
we use an fc layer to regress to the bounding box offsets for each object class separately .
we use cross - entropy to compute the loss on the object class and the relationship predicate .
we use l1 function to compute the loss of the bounding box offsets .
we use an ms coco - pretrained vgg-16 network to extract visual features from images .
we freeze the weights of all convolution layers , and only finetune the fully connected layers , including the grus .
the node grus and the edge grus have both 512-dimensional input and output . during training
, we first use nms to select at most 2,000 boxes from all proposed boxes @xmath0 , and then randomly select 256 boxes as the object proposals . due to the quadratic number of edges , we randomly sub - sample 512 edges for each image at training time .
edges without predicate labels are assigned to `` none '' class . at test time
, we use nms to select at most 50 boxes from the object proposals with an iou threshold of 0.3 .
we make predictions on all edges except the self - connections at the test time .
we evaluate our model on generating scene graphs from images .
we compare our model against a recently proposed model on visual relationship prediction @xcite .
our goal is to analyze our model in datasets with both sparse and dense relationship annotations .
we use the visual genome scene graph dataset @xcite in our main experiment .
we also evaluate our model on the support relation inference task in the nyu depth v2 dataset .
the key difference between these two datasets is that scene graph annotation is very sparse in visual genome : among all possible pairing of objects , only 5% of them are labeled with a relationship predicate .
the nyu depth v2 dataset , on the other hand , exhaustively annotates the support of every labeled object .
our experiments show that our model outperforms the baseline model @xcite , and can generalize to other types of relationships , in particular support relations @xcite , without any architecture change .
* visual genome * the visual genome dataset @xcite is a super set of the visual relationship dataset used in @xcite .
it contains 108,077 images annotated with on average 13.5 objects and 15 relationships per image .
the object categories and relationship predicates are annotated with an open vocabulary .
there are 75,729 unique object categories , and 40,480 unique relationship predicates .
as the annotations follow an extremely long - tail distribution , we use the most frequent 150 object categories and 50 predicates for evaluation . as a result
, each image has a scene graph of around 12 objects and 7 relationships .
we use 70% of the images for training and the remaining 30% for testing . *
nyu depth v2 * we also evaluate our model on the support relation graphs from the nyu depth v2 dataset @xcite .
the dataset contains 1,449 rgbd images captured in 27 indoor scenes .
each image is annotated with instance segmentation , region class labels , and support relations between regions .
we use the standard split , with 795 images used for training and 654 images for testing .
[ [ setup ] ] setup + + + + + given an image , the scene graph generation task is to localize a set of objects , classify their category labels , and predict relationships between each pair of the objects .
we evaluate our model on the visual genome dataset .
we analyze our model in three setups below . 1 . the * predicate classification * ( predcls ) task is to predict the predicates of all pairwise relationships of a set of localized objects .
this task examines the model s performance on predicate classification in isolation from other factors . 2 . the * scene graph classification * ( sgcls ) task is to predict the predicate as well as the object categories of the subject and the object in every pairwise relationship given a set of localized objects .
the * scene graph generation * ( sggen ) task is to simultaneously detect a set of objects and predict the predicate between each pair of the detected objects .
an object is considered to be correctly detected if it has at least 0.5 iou overlap with the ground - truth box .
we adopted the image - wise recall evaluation metrics , r@50 and r@100 , that are used in lu @xcite for all the three setups .
the r@@xmath39 metric measures the fraction of ground - truth relationship triplets ` ( subject- predicate - object ) ` that appear among the top @xmath39 most confident triplet predictions in an image .
the choice of this metric is , as explained in @xcite , due to the sparsity of the relationship annotations in visual genome metrics like map would falsely penalize positive predictions on unlabeled relationships .
we also report per - type recall@5 of classifying individual predicate .
this metric measures the fraction of the time the correct predicate is among the top 5 most confident predictions of each labeled relationship triplet .
as shown in table [ table : pred_type ] , many predicates have very similar semantic meanings , for example , ` on ` vs. ` over ` and ` hanging from ` vs. ` attached to ` .
the less frequent predicates would be overshadowed by the more frequent ones during training .
we use the recall metric to alleviate such an effect . [
fig : train_iter ] we evaluate our final model and a number of baseline models .
one of the key components in our primal - dual formulation is the message pooling functions that use learnt weighted sum to aggregate hidden states of nodes and edges into messages ( see eq .
[ eq : vert ] and eq .
[ eq : edge ] ) . in order to demonstrate its effectiveness ,
we evaluate variants of our model with standard pooling methods .
the first is to use average - pooling ( * avg .
pool * ) instead of the learnt weighted sum to aggregate the hidden states .
the second is similar to the first one , but uses max - pooling ( * max pool * ) .
we also evaluate our models against a relationship detection model proposed by lu @xcite .
their model consists of two components a vision module that makes predictions from images , and a language module that captures language priors .
we compare with their vision module , which uses the same inputs as ours ; their language module is orthogonal to our model , and can be added independently .
note that this model is equivalent to our final model without any message passing .
.evaluation results of the scene graph generation task on the visual genome dataset @xcite .
we compare a few variations of our model against a visual relationship detection module proposed by lu @xcite ( sec .
[ sec : models ] ) . [ cols="<,<,^,^,^,^,^",options="header " , ] [ table : nyu_eval ] we then evaluate on the nyu depth v2 dataset @xcite with densely labeled support relations .
we show that our model can generalize to other type of relationships and is effective on both sparsely and densely labeled relationships .
[ [ setup-1 ] ] setup + + + + + the nyu depth v2 dataset contains three types of support relationships : an object can be supported by an object from behind , by an object from below , or supported by a hidden object .
each object is also labeled with one of the four structure classes : ` { floor , structure , furniture , prop } ` .
we define the support graph generation task as to predicting both th support relation type between objects and the structure class of each object .
we take the smallest bounding box that encloses an object segmentation mask as its object region .
we assume ground - truth object locations in this task .
we compare our final model with two previous models @xcite on the support graph generation task . following the metric used in previous work , we report two types of support relation accuracies @xcite : type - aware and type - agnostic .
we also report the performance with r@50 and r@100 measurements of the predicate classification task introduced in sec .
[ sec : vg_eval ] .
note that both @xcite and @xcite use rgbd images , whereas our model uses only rgb images .
[ [ results-1 ] ] results + + + + + + + our model outperforms previous work , achieving new state - of - the - art performance using only rgb images .
our results show that having contextual information further improves support relation prediction , even compared to purpose - built models @xcite that used rgbd images .
[ fig : nyu ] shows some sample predictions using our final model .
incorrect predictions typically occur in ambiguous supports , e.g. , books in shelves can be mistaken as being supported from behind ( row 1 , column 2 ) .
another failure mode is due to geometric structures that have weak visual features .
as shown in row 2 , column 1 , the ceiling at the top left corner of the image is predicted as supported from behind instead of supported below by the wall , but the boundary between the ceiling and the wall is nearly invisible .
such visual uncertainty may be resolved by having additional depth information .
: support from below , @xmath40 : support from behind .
red arrows are incorrect predictions .
we also color code structure classes : ground is in blue , structure is in green , furniture is in yellow , prop is in red .
purple indicates missing structure class .
note that the segmentation masks are only shown for visualization purpose .
our model uses object bounding boxes as input instead of these masks ]
we addressed the problem of automatically generating a visually grounded scene graph from an image by a novel end - to - end model .
our model performs iterative message passing between the primal and dual sub - graph along the topological structure of a scene graph . this way , it improves the quality of node and edge predictions by incorporating informative contextual cues .
our model can be considered a more generic framework for graph generation problem . in this work
, we have demonstrated its effectiveness in predicting visual genome scene graphs as well as support relations in indoor scenes .
a possible future direction would be to explore its capability in other structured prediction problems in vision and other problem domains .
we would like to thank ranjay krishna , judy hoffman , junyoung gwak , and anonymous reviewers for useful comments .
this research is partially supported by a yahoo labs macro award , and an onr muri award . c. desai , d. ramanan , and c. fowlkes .
discriminative models for static human - object interactions . in _
2010 ieee computer society conference on computer vision and pattern recognition - workshops_. ieee , 2010 . c. galleguillos , a. rabinovich , and s. belongie .
object categorization using co - occurrence , location and appearance . in _ computer vision and pattern recognition , 2008 .
cvpr 2008 .
ieee conference on_. ieee , 2008 .
j. johnson , r. krishna , m. stark , l. j. li , d. a. shamma , m. s. bernstein , and l. fei - fei .
image retrieval using scene graphs . in _
ieee conference on computer vision and pattern recognition ( cvpr ) _ , 2015 .
r. krishna , y. zhu , o. groth , j. johnson , k. hata , j. kravitz , s. chen , y. kalantidis , l .- j .
li , d. a. shamma , m. bernstein , and l. fei - fei .
visual genome : connecting language and vision using crowdsourced dense image annotations . in _ arxiv _ , 2016 .
d. lin , s. fidler , and r. urtasun .
holistic scene understanding for 3d object detection with rgbd cameras . in _ proceedings of the ieee international conference on computer vision _ , pages 14171424 , 2013 .
v. ramanathan , c. li , j. deng , w. han , z. li , k. gu , y. song , s. bengio , c. rossenberg , and l. fei - fei .
learning semantic relationships for better action retrieval in images . in _
2015 ieee conference on computer vision and pattern recognition ( cvpr)_. ieee , 2015 .
r. salakhutdinov , a. torralba , and j. tenenbaum .
learning to share visual appearance for multiclass object detection . in _
computer vision and pattern recognition ( cvpr ) , 2011 ieee conference on_. ieee , 2011 .
b. yao and l. fei - fei .
modeling mutual context of object and human pose in human - object interaction activities . in _
computer vision and pattern recognition ( cvpr ) , 2010 ieee conference on_. ieee , 2010 .
s. zheng , s. jayasumana , b. romera - paredes , v. vineet , z. su , d. du , c. huang , and p. torr .
conditional random fields as recurrent neural networks . in _ international conference on computer vision ( iccv ) _ , 2015 . | understanding a visual scene goes beyond recognizing individual objects in isolation .
relationships between objects also constitute rich semantic information about the scene . in this work , we explicitly model the objects and their relationships using scene graphs , a visually - grounded graphical structure of an image .
we propose a novel end - to - end model that generates such structured scene representation from an input image .
the model solves the scene graph inference problem using standard rnns and learns to iteratively improves its predictions via message passing .
our joint inference model can take advantage of contextual cues to make better predictions on objects and their relationships .
the experiments show that our model significantly outperforms previous methods on generating scene graphs using visual genome dataset and inferring support relations with nyu depth v2 dataset . |
flow inside a porous medium finds many applications in natural and engineering systems .
subsurface flows @xcite , erosion @xcite , fuel cells @xcite , and filtration systems @xcite are a few examples of physical processes that are governed by low reynolds number porous media flow . accurately resolving such flows is essential for modelling transport @xcite , mixing @xcite , anomalous diffusion @xcite , breakthrough curves @xcite , carbon dioxide sequestration @xcite , uncertainty quantification @xcite , chemical reactions @xcite , and many other applications . in most of these works , in particular , those that focus on numerics and modelling , the porous medium is assumed to be two dimensional .
therefore , there is a strong need for efficient and accurate numerical methods to simulate complex two - dimensional porous media flow . in this paper , we focus on solving the two - dimensional incompressible stokes equations in the geometry illustrated in .
the incompressible stokes equations have been solved in porous domains resembling using finite differences @xcite , finite volumes @xcite , algebraic multigrid @xcite , lattice - boltzmann @xcite , and continuous time random walks @xcite .
however , each of these methods struggle with either enforcing incompressibility , meshing the geometry , or obtaining high - order accuracy .
boundary integral equations ( bies ) , on the other hand , are a powerful method that addresses each of these challenges .
in particular , bies are naturally adaptive , allow for high - order approximations , automatically enforce the incompressibility constraint , and enable the computation of the velocity field without computing the pressure .
another advantage of a bie is that the unknown two - dimensional velocity field is written in terms of a one - dimensional integral over the boundary of the domain .
this dimension reduction implies that only the boundary , which is a collection of one - dimensional closed curves , needs to be meshed .
furthermore , a natural extension of this work is to solve the three - dimensional stokes equations while only discretizing the two - dimensional boundary of the geometry .
other groups have used integral equations to simulate porous media by using a darcy approximation @xcite , or by coupling a penalization method with a volume integral equation @xcite .
in contrast , we will use a boundary integral equation to solve the incompressible stokes equations .
the main disadvantage of bies is the need to solve a dense linear system of equations . a direct solver based on gaussian elimination requires @xmath3 operations where @xmath4 denotes the number of unknowns which is impractical for realistic geometries .
it also requires the assembly and storage of the dense matrix , leading to an @xmath5 memory requirement .
iterative solvers such as gmres , on the other hand , only rely on ( accelerated ) matrix - vector multiplications and have an @xmath6 memory requirement , but can suffer from a poor convergence behavior due to the ill - conditioning of the linear system .
in the recent decade , there have been two main directions of research that have lead to faster algorithms for solving discretized bies .
a first approach focuses on developing integral equations that result in linear systems that are both well - conditioned and require a mesh - independent number of gmres iterations to reach convergence @xcite .
these properties generally follow from second - kind integral equation methods with compact integral operators . because of the mesh - independent number of gmres iterations , the asymptotic cpu time scales with the cost of a single dense matrix - vector multiplication . using fast summation methods such as the fast multipole method ( fmm ) @xcite , ewald summation @xcite , or tree - code methods @xcite
, the cost of a matrix - vector multiplication can be reduced from @xmath5 to either @xmath7 or @xmath6 . while this approach can result in an optimal asymptotic complexity , complexities of the geometry , such as nearly touching pores and regions of high curvature , often still results in a large number of gmres iterations @xcite . in addition , the double - layer potential formulation , which requires a mesh - independent number of gmres iterations , has less regularity than the single - layer potential @xcite , which requires a mesh - dependent number of gmres iterations . to reduce the number of iterations
, a preconditioner can be incorporated .
while simple techniques such as block - diagonal preconditioners are easy to construct , they are not always capable of reducing the number of iterations significantly .
an incomplete list of more sophisticated preconditioners , which vary in construction time and efficacy , include additive or multiplicative schwarz @xcite , bpx @xcite , sparse approximate inverses @xcite , approximate lu decompositions @xcite , clustering @xcite , integral equations of opposite order @xcite , calderon identities @xcite , or multigrid @xcite . in the case of multigrid preconditioners ,
specialized smoothers must be developed since standard smoothers like jacobi or gauss - seidel will smooth the low , rather than the high , frequency components .
examples of such smoothers may require developing approximate inverses of the integral operator @xcite involving , for example , the laplace - beltrami operator @xcite .
a second approach targets the use of fast direct solvers that exploit the structure of the linear system that arises when discretizing bies .
the structure can be loosely described as matrices with hierarchical off - diagonal blocks being numerically low - rank .
these matrices are formalized as , with an increasing level of complexity , hierarchically off - diagonal low - rank ( hodlr ) matrices @xcite , hierarchically semi - separable ( hss ) matrices @xcite , @xmath8-matrices @xcite , and @xmath0-matrices @xcite .
fast methods for efficiently constructing and storing approximate inverses of these matrices have been developed in recent years @xcite , mainly for the aforementioned hodlr and hss matrices .
moreover , once the approximate inverse , which only depends on the geometry , is computed and stored , it can be applied with a minimal amount of cpu time to multiple right - hand sides .
therefore , direct solvers are extremely useful when applying a temporal discretization to a partial differential equation @xcite , or when considering scattering for many different incident waves @xcite .
we blend the two aforementioned approaches in this paper : we present the inverse fast multipole method ( ifmm ) @xcite as an inexact fast direct solver based on @xmath0-matrices and apply it as a preconditioner in a fast multipole accelerated gmres solver @xcite .
the preconditioner has a tunable accuracy @xmath1 and its computational cost scales almost linearly with the problem size as @xmath9 .
we will demonstrate that the ifmm is capable of significantly reducing the number of iterations and the cpu time .
the outline of this paper is as follows .
discusses the fluid model and layer potential formulation
. the inverse fast multipole method ( ifmm ) is subsequently introduced in , focusing on the main features for achieving an efficient preconditioner .
numerical benchmarks are presented in .
we report results for both the unpreconditioned system , and for preconditioned systems where block - diagonal and ifmm preconditioners ( at several accuracies ) are used . summarizing conclusions
are finally drawn in .
we are interested in simulating an incompressible newtonian fluid in the geometry illustrated in .
we consider scales where the dimensionless reynolds number is small so that the fluid is assumed to be stokesian . if we assumed a homogeneous porosity , the much simpler darcy equations could be solved .
however , we do not make this assumption and we focus on accurately and efficiently solving the incompressible stokes equation with a no - slip boundary condition on each pore .
we also assume that the flow is two - dimensional and that there are no external body forces such as gravity .
we avoid the stokes paradox by bounding the porous region by the boundary @xmath10 .
to resemble a pipe flow , we let @xmath10 be a mollification of the rectangle @xmath11 \times [ -h , h]$ ] ( see ) . by smoothing the corners of @xmath10 , specialized quadrature for integral equations
is avoided .
each inner boundary , denoted by @xmath12 , @xmath13 , is a circle of variable radii .
the area enclosed by the outer boundary is denoted by @xmath14 , and the area enclosed by each pore by @xmath15 .
therefore , the geometry is given by @xmath16 and its boundary is @xmath17 . with this setup , by defining the fluid velocity @xmath18 , the pressure @xmath19 , the viscosity @xmath20 , and a prescribed flow on the boundary @xmath21 , the governing equations are @xmath22 we impose boundary conditions that correspond to a pipe flow at the intake and outtake , and no slip on the boundary of the pores @xmath23 where @xmath24 sets the velocity scale .
furthermore , we assume that @xmath25 from this point onwards .
a bie formulation of has several advantages over the differential form .
the pressure need not be computed , the incompressibility constraint is automatically satisfied , and the velocity is written in terms of a layer potential involving an unknown density function defined only on @xmath26 .
two popular choices for the layer potential are the single - layer and double - layer potentials @xmath27({\mathbf{x } } ) & = \frac{1}{4\pi } \int_{\gamma } \left ( -\log \rho \mathbf{i } + \frac{{\mathbf{r}}\otimes { \mathbf{r}}}{\rho^{2}}\right ) { \boldsymbol{\sigma}}({\mathbf{y } } ) ds_{{\mathbf{y } } } , \qquad & & { \mathbf{x}}\in \omega , \\ \mathcal{d}[{\boldsymbol{\sigma}}]({\mathbf{x } } ) & = \frac{1}{\pi}\int_{\gamma } \frac{{\mathbf{r}}\cdot { \mathbf{n}}}{\rho^{2 } } \frac{{\mathbf{r}}\otimes { \mathbf{r}}}{\rho^{2 } } { \boldsymbol{\sigma}}({\mathbf{y } } ) ds_{{\mathbf{y } } } , & & { \mathbf{x}}\in \omega,\end{aligned}\ ] ] respectively , where @xmath28 , @xmath29 , @xmath30 is the unit outward normal , and @xmath31 is a density function defined on @xmath26 . given our dirichlet boundary condition , the double - layer potential formulation results in a second - kind integral equation whose condition number is guaranteed to be mesh - independent .
however , the single - layer potential requires less smoothness of @xmath31 since @xcite @xmath32 are continuous .
moreover , this result holds even under the relaxed assumption that @xmath33 is a bounded lipschitz domain .
another advantage of the single - layer potential is that it has full - rank , as opposed to the double - layer potential which has a rank three null space @xcite for each connected component of @xmath26 .
while this rank deficiency can be removed , it is unclear if this is compatible with the ifmm .
moreover , for complex geometries , the double - layer potential may still require many gmres iterations @xcite , but in this case , we expect that applying the ifmm preconditioner will result in a computational speedup .
given the additional regularity and full - rank of the single - layer potential , we choose to represent the velocity @xmath34 as @xmath35({\mathbf{x } } ) = \frac{1}{4\pi } \int_{\gamma } \left ( -\log \rho \mathbf{i } + \frac{{\mathbf{r}}\otimes { \mathbf{r}}}{\rho^{2}}\right ) { \boldsymbol{\sigma}}({\mathbf{y } } ) ds_{{\mathbf{y } } } , \quad { \mathbf{x}}\in \omega , \label{eq : stokesslp}\end{aligned}\ ] ] where @xmath31 is an unknown density function .
taking the limiting value of as @xmath36 tends to @xmath26 , the density function @xmath31 must satisfy the first - kind integral equation @xcite @xmath37 it is known that upon discretizing , the result is an ill - conditioned linear system that requires a mesh - dependent number of gmres iterations .
however , the linear system is amenable to the solvers for structured matrices .
once the unique solution of is computed , the density function @xmath31 is substituted into to evaluate the velocity @xmath18 at any point @xmath38 .
we adopt a collocation method to discretize .
first , each interior curve @xmath12 , @xmath13 is parameterized and discretized at @xmath39 points , and the outer wall is discretized at @xmath40 points .
the resulting discretization points are @xmath41 , @xmath42 .
we enforce the bie at these @xmath43 collocation points by requiring @xmath44 then , quadrature is applied , where the quadrature nodes coincide with the discretization points @xmath41 .
this results in the dense linear system @xmath45 where @xmath46 is the kernel of the integral operator in .
the weights @xmath47 are the product of the arclength term with appropriate quadrature weights .
we integrate the weak logarithmic singularity by applying a sixth - order quadrature rule of kapur and rokhlin @xcite .
this quadrature rule is identical to the trapezoid rule , except that the weights are modified at the six nodes to the left and right of the singularity .
therefore , the quadrature nodes coincide with the discretization points @xmath41 , @xmath48 . since there are only @xmath49 quadrature weights that are modified for each @xmath41 , these quadrature rules are compatible with the fmm .
in addition , in contrast to alpert quadrature rules @xcite , additional quadrature nodes are not introduced , so no interpolation is required .
given this quadrature formula , results in a dense @xmath50 linear system that , for brevity , we write as @xmath51 the matrix @xmath52 is a structured matrix with numerical low - rank off - diagonal blocks .
gmres can be used to iteratively solve , and each matrix - vector multiplication can be done with @xmath6 complexity by using the fmm . however , the number of required gmres iterations will be large , and depend on the mesh size .
therefore , we will use the structure of @xmath52 to develop an efficient preconditioner for gmres .
upon computing the density function @xmath31 at the points @xmath53 , the velocity field needs to be computed at points @xmath38 .
this can be done with the trapezoid rule @xmath54 where @xmath55 is an arclength spacing of @xmath26 at @xmath56 .
the trapezoid rule guarantees spectral accuracy , but this asymptotic convergence rate is delayed when @xmath36 is too close to @xmath26 . ] .
there are methods to guarantee a uniform error for all @xmath57 @xcite , but this is not the focus of this work .
instead , we focus on efficiently solving for the density function @xmath31 .
the inverse fast multipole method ( ifmm ) @xcite is an inexact fast direct solver for @xmath0-matrices that can be applied as a preconditioner for gmres .
it has successfully been used to precondition matrices arising from integral operators @xcite and radial basis interpolation @xcite .
the main features of the algorithm are concisely described in the following subsections ; the reader is referred to @xcite for a more detailed description . although a 2d problem is considered in this paper ,
the ifmm is applicable to 3d problems as well ( see @xcite for examples ) .
a variety of fast direct solvers for dense linear systems such as have been developed in recent years .
these methods are based on a multilevel , hierarchical decomposition of the physical domain of interest , from the root down to the leaf level @xmath58 ( e.g. , by means of a quadtree in 2d or an octree in 3d , either uniform or adaptive ) .
this is subsequently exploited to represent the matrix @xmath52 in a compressed format .
the simplest of these methods assume that all off - diagonal blocks are low - rank ( only self - interactions are considered to be of full - rank ) , leading to so - called hierarchically off - diagonal low - rank ( hodlr ) @xcite and hierarchically semi - separable ( hss ) matrices @xcite . in the latter ,
a nested approach is used ( i.e. , the low - rank basis at a certain hierarchical level is constructed using the low - rank basis at the child level ) , while this is not the case in the former .
given this assumption on the matrix structure , the inversion can be performed exactly .
direct @xmath6 algorithms for hodlr matrices can be found in @xcite , while fast solvers for hss matrices have been presented by , among others , martinsson and rokhlin @xcite , chandrasekaran et al .
@xcite , xia et al .
@xcite , and gillman et al . @xcite .
the assumption that all off - diagonal blocks are low - rank is a major drawback for the aforementioned hodlr and hss matrices .
indeed , the rank is in that case likely to grow in an unbounded fashion for 2d and 3d problems with a complex geometry @xcite , hence jeopardizing the computational efficiency of the algorithms .
this drawback can be circumvented by adopting the more general framework of @xmath8- or @xmath0-matrices ( non - nested vs. nested ) , in which only non - neighboring interactions are replaced by low - rank approximations .
this is the approach followed in the ifmm @xcite .
it is worth noting that the algorithm presented in @xcite has recently been extended by corona et al .
@xcite to obtain a direct solver for hss matrices that scales for 2d problems as @xmath6 .
we note that since our problem is two - dimensional , a hss approximation of would potentially have bounded rank . however , to extend the work to three dimensions , we choose to use an @xmath0 framework .
an @xmath0-matrix @xmath59 is weak hierarchical matrix ( in which only non - neighboring , well - separated interactions are assumed to be low - rank ) with a nested basis that approximates the original matrix @xmath52 .
it can be expressed as @xcite @xmath60 @xmath59 is thus decomposed into a block sparse matrix , @xmath61 , containing all the neighboring and self - interactions at the leaf level @xmath58 , and a low - rank term , @xmath62 , that characterizes all well - separated interactions . @xmath63 and @xmath64 are interpolation and anterpolation operators , respectively , while @xmath65 is an @xmath0-matrix of a smaller size .
is applied recursively to @xmath65 until the top level . ] , where there are no more well - separated interactions , is reached .
the ifmm is a novel algorithm for the approximate inversion of an @xmath0-matrix , and consists of two crucial steps : extended sparsification and compression of fill - ins that appear throughout the elimination .
the hierarchical structure of @xmath59 can be exploited to represent @xmath66 as an extended sparse system , which is more attractive from a computational point of view . introducing auxiliary variables @xmath67 and @xmath68 ( corresponding to multipole and local coefficients in the fmm ) leads to the following extended system : @xmath69 in , the sparse blocks @xmath61 , @xmath63 , and @xmath70 are situated in the top left corner of the matrix , while the remaining dense block @xmath65 is pushed to the bottom right corner of the matrix . as the latter is an @xmath0-matrix as well , the same technique can be applied recursively to further extend and sparsify the system .
one finally obtains the following system of equations , which is equivalent to @xmath66 : @xmath71 the variables and equations are ordered in such a way that a sparse matrix with a symmetric fill - in pattern is obtained .
note that this extended system is only moderately larger than the original system , since the dimensions of the auxiliary variables ( characterizing the low - rank interactions ) are small compared to the dimension of @xmath31 .
the application of gaussian elimination to the extended sparse system of leads to the creation of dense fill - in blocks , which jeopardizes the computational efficiency of the method . in order to obtain a fast solver
, it is crucial to preserve the sparsity pattern throughout the elimination .
this can be achieved by compressing and redirecting fill - ins that correspond to well - separated interactions , as these are expected to have a numerically low - rank .
for example , consider the fill - ins @xmath72 and @xmath73 arising between the well - separated variables @xmath74 and @xmath75 ( the prime indicates a fill - in ; @xmath76 and @xmath77 are originally zero blocks ) .
these fill - ins can be approximated by a low - rank representation ( compression step ) by means of a truncated singular value decomposition ( svd ) in which only the most significant singular values and vectors are retained : @xmath78 the rank of approximations and is determined by a prescribed tolerance @xmath1 ( either relative or absolute ) on the singular values .
next , a recompression step is performed to obtain new low - rank interpolation and anterpolation operators @xmath79 and @xmath80 , respectively , as well as @xmath81 and @xmath82 , such that the fill - ins @xmath72 and @xmath73 can be redirected through the existing low - rank interaction between @xmath74 and @xmath75 in @xmath65 . as a result
, there is no need to store @xmath72 and @xmath73 explicitly , implying that the sparsity pattern of the matrix is maintained throughout the elimination .
if we make the assumption that the rank remains bounded as the problem size increases , it can be demonstrated that this leads to an algorithm with an asymptotic complexity of @xmath9 .
although there is no mathematical proof to support this assumption , numerical examples confirm the fact that this is indeed the case in many practical examples @xcite .
the compression and recompression procedures make the ifmm inexact , but the accuracy can be tuned by varying @xmath1 . in terms of computational efficiency , one needs to make a trade - off between a highly accurate direct solver or a low - accuracy preconditioner .
we focus in this paper on its use as a preconditioner .
more details on the aforementioned procedures are provided in @xcite .
the value of @xmath1 should be chosen in relation to the accuracy of the initial low - rank operators in ( i.e. , the higher the initial rank , the smaller @xmath1 should be ) .
several numerical benchmarks are considered in this section to demonstrate the efficiency of the ifmm as a preconditioner for the problems under concern . in all of these benchmarks , is solved for the density vector @xmath83 with a fast multipole accelerated gmres solver @xcite , and is subsequently used to evaluate the velocities @xmath84 inside the domain .
a tolerance of @xmath85 is specified for the relative residual in gmres .
the computations have been performed on intel^^ xeon^^ e5 - 2650 v2 ( 2.60 ghz ) cpus . in all benchmarks ,
a black - box approach is used in the ifmm for constructing the initial low - rank operators in .
more precisely , interpolation based on chebyshev polynomials @xcite is employed ( with @xmath86 chebyshev nodes in each direction ) , followed by an additional svd to reduce the rank .
a uniform quadtree decomposition of the domain is used , and the number of levels is adjusted to provide a reasonable trade - off between the time spent at the leaf level and the time spent at higher levels in the tree .
we are interested in the porous geometry illustrated in .
the outer wall is a rounded off version of the rectangle @xmath87 \times [ -2.6,2.6]$ ] , and the radii of the pores are distributed in @xmath88 $ ] .
the smallest gap between two pores is @xmath89 , and the smallest gap between a pore and the outer wall is @xmath90 .
finally , the porous region has a porosity of @xmath91 .
we test the preconditioner on simpler geometries by considering only the first @xmath92 and @xmath93 left - most pores .
we also scale the outer boundary so that the area to the right of the pores remains roughly constant . before studying the pipe flow boundary conditions defined in , we first validate our method by examining the convergence with two sets of boundary conditions that have a closed - form solution in and [ subsec : analytical ] .
we subsequently discuss the pipe flow boundary conditions of in .
we use a shear flow boundary condition on all the boundaries so that the exact solution is @xmath94 . with this boundary condition ,
the density function @xmath83 is only non - zero along the outer boundary @xmath10 .
a first small example involves only the first 22 pores ( see ) ; the geometry of the outer boundary @xmath96 is scaled so that the area of the pore - free region is the same as the area of the pore - free region in .
we discretize each pore with @xmath97 points and the outer boundary with @xmath98 points .
the resulting dense system has nearly @xmath99 unknowns .
three different preconditioning approaches are used in gmres : ( i ) no preconditioner , ( ii ) a block - diagonal preconditioner ( with the diagonal blocks corresponding to the self - interactions of the individual pores and the outer boundary ) , and ( iii ) the ifmm . in the ifmm , ten chebyshev nodes are employed in each direction to construct the initial low - rank approximations ( @xmath100 ) , along with a relative accuracy @xmath101 for the low - rank ( re)compressions . the preconditioners in cases ( ii ) and ( iii ) are applied from the left as @xmath102 where @xmath103 represents the preconditioner .
\(a ) shows the relative residual as a function of the number of gmres iterations and ( b ) shows the cpu time required by the preconditioner and gmres for all three preconditioning strategies .
convergence is very slow without a preconditioner and the relative residual does not drop below the desired tolerance within the prescribed maximum number of 1000 iterations .
using the block - diagonal preconditioner , on the other hand , leads to convergence within 139 iterations .
this is a very cheap preconditioning strategy , as the time needed to construct the preconditioner is negligible .
the ifmm preconditioner is more expensive to compute , but the overall wall time is significantly reduced because the total number of iterations is reduced to merely 40 .
note that the relative residual shown in ( a ) is the left preconditioned residual @xmath104 , where @xmath105 is the estimated solution at each gmres iteration .
the actual residual @xmath106 is computed as well once the latter drops below the prescribed tolerance .
indicates that the discrepancy between the preconditioned and the actual relative residual is very small in all cases .
the overall computation times for each of the preconditioning approaches are presented in ( b ) and , where a decomposition is made into the time required to construct the preconditioner and the actual iteration time .
it is clear that the additional cost for constructing the ifmm pays off since the total time is more than halved with respect to the block - diagonal preconditioner .
( a ) pores and @xmath107 unknowns . the number of gmres iterations is restricted to 1000.,title="fig : " ] ( b ) pores and @xmath107 unknowns .
the number of gmres iterations is restricted to 1000.,title="fig : " ] .overview of the results for various preconditioning strategies for a shear flow problem with @xmath95 pores and @xmath107 unknowns .
the number of gmres iterations is restricted to 1000 .
[ cols="^,^,^,^,^ " , ] shows a quiver plot of the velocity field @xmath18 obtained with the ifmm preconditioner .
a very complex flow is resolved in this case as well .
obtained when using the ifmm preconditioner for a pipe flow problem with @xmath108 pores and @xmath109 unknowns .
the velocity field is plotted in a small window in the middle of the porous region . ]
discretization of a first - kind boundary integral equation , representing a 2d incompressible stokes flow in a porous medium , leads to a dense linear system of equations that is computationally expensive to solve .
in particular , a large number of iterations is often required if an iterative krylov methods such as gmres is employed . in this paper ,
the ifmm has been presented as an efficient preconditioner that is capable of significantly reducing the number of iterations and the overall computation cost .
the ifmm is in essence an inexact fast direct solver for dense @xmath0-matrices with a linear complexity .
this complexity is achieved through two key ideas : the hierarchical low - rank structure of the matrix is exploited to represent the original system as an extended sparse system , and low - rank ( re)compressions are applied to maintain the sparsity pattern of the latter throughout the gaussian elimination .
the solver has a tunable accuracy @xmath1 allowing for a trade - off between a highly accurate direct solver and a low - accurate preconditioner .
various numerical benchmarks have been carried out to validate the accuracy and to assess the efficacy of the ifmm as a preconditioner .
it has been demonstrated that the ifmm preconditioner outperforms a block - diagonal preconditioner for several types of boundary conditions .
this is especially the case for pipe flow problems involving many pores , which is relevant for various practical applications . finally , it is important to note that the current ifmm implementation is based on a uniform quadtree decomposition of the domain , which is far from optimal for the complex geometries considered in this paper .
the use of an adaptive quadtree is expected to make the ifmm even more efficient .
pc is a post - doctoral fellow of the research foundation flanders ( fwo ) and a francqui foundation fellow of the belgian american educational foundation ( baef ) . the financial support is gratefully acknowledged .
bq acknowledges startup funds from florida state university .
kitamura k , jiang f , valocchi aj , chiyonobu s , tsuji t , christensen kt .
the study of heterogeneous two - phase flow around small - scale heterogeneity in porous sandstone by measured elastic wave velocities and lattice boltzmann method simulation .
_ journal of geophysical research : solid earth _ 2014 ; * 119*(10):75647577 .
wu d , liu h , xie m , liu h , sun w. experimental investigation on low velocity filtration combustion in porous packed bed using gaseous and liquid fuels . _ experimental thermal and fluid science _ 2012 ; * 36*:169177 .
kazemifar f , blois g , kyritsis dc , christensen kt . quantifying the flow dynamics of supercritical co2water displacement in a 2d porous micromodel using fluorescent microscopy and microscopic piv .
_ advances in water resources _ 2015 ; .
lebon l , oger l , leblond j , hulin jp , martys ns , schwartz lm .
pulsed gradient nmr measurements and numerical simulation of flow velocity distribution in sphere packings .
_ physics of fluids _ 1996 ; * 8*(2):293301 .
de anna p , borgne tl , dentz m , tartakovsky am , bolster d , davy p. flow intermittency , dispersion , and correlated continuous time random walks in porous media . _ physical review letters _ 2013 ; * 110*(18):184502 .
malholtra d , gholami a , biros g. a volume integral equation stokes solver for problems with variable coefficients . _
proceedings of the international conference for high performance computing , networking , storage and analysis _
, ieee press , 2014 ; 92102 .
heuer n , stephan ep , tran t. multilevel additive schwarz method for the h - p version of the galerkin boundary element method .
_ mathematics of computation of the american mathematical society _ 1998 ; * 67*(222):501518 .
nabors k , korsmeyer ft , leighton ft , white jk .
preconditioned , adaptive , multipole - accelerated iterative methods for three - dimensional first - kind integral equations of potential theory .
_ siam journal on scientific and statistical computing _ 1994 ; * 15*:713735 .
langer u , pusch d , reitzinger s. efficient preconditioners for boundary element matrices based on grey - box algebraic multigrid methods .
_ international journal for numerical methods in engineering _ 2003 ; * 58*(13):19371953 .
sheng z , dewilde p , chandrasekaran s. algorithms to solve hierarchically semi - separable systems .
_ system theory , the schur algorithm and multidimensional analysis _ , _ operator theory : advances and applications _ ,
176 , alpay d , vinnikov v ( eds . ) .
birkhuser basel , 2007 ; 255294 .
gillman a , martinsson pg .
an @xmath113 algorithm for constructing the solution operator to 2d elliptic boundary value problems in the absence of body loads .
_ advances in computational mathematics _ 2014 ; * 40*(4):773796 .
coulier p , darve e. efficient mesh deformation based on radial basis function interpolation by means of the inverse fast multipole method .
_ computer methods in applied mechanics and engineering _ 2016 ; . | we consider an efficient preconditioner for boundary integral equation ( bie ) formulations of the two - dimensional stokes equations in porous media . while bies are well - suited for resolving the complex porous geometry , they lead to a dense linear system of equations that is computationally expensive to solve for large problems .
this expense is further amplified when a significant number of iterations is required in an iterative krylov solver such as gmres . in this paper
, we apply a fast inexact direct solver , the inverse fast multipole method ( ifmm ) , as an efficient preconditioner for gmres .
this solver is based on the framework of @xmath0-matrices and uses low - rank compressions to approximate certain matrix blocks .
it has a tunable accuracy @xmath1 and a computational cost that scales as @xmath2 .
we discuss various numerical benchmarks that validate the accuracy and confirm the efficiency of the proposed method .
we demonstrate with several types of boundary conditions that the preconditioner is capable of significantly accelerating the convergence of gmres when compared to a simple block - diagonal preconditioner , especially for pipe flow problems involving many pores . |
game theory has experienced a great development since early time of last century through its wide application in nearly every aspect of our modern society . in recent few years , along with the thriving of quantum information , a newly - emerging field , quantum game has begun to attract much attention due to the intimate connection between the theory of game and the theory of quantum communication @xcite .
this field was initiated by a paper of meyer , finding that a player can always beats his classical opponent by adopting quantum strategies on the coin tossing game @xcite .
later , eisert _ et al . _
@xcite introduced quantum entanglement into the famous prisoners dilemma , by which the contradiction existing in the classical scenario vanishes . since then
, quantum game theory has developed rapidly and exhibited great superiority over its classical counterpart in later works @xcite .
in addition to theoretical investigations , du _
realized the quantum prisoners dilemma in experiments for the first time @xcite , and therefore constructed a practical base for quantum game . however , most of previous investigations on quantum game mainly focused on game models with discrete strategies set . in li _
_ s paper @xcite , they first put forward a quantization scheme for cournot duopoly , a famous economical model even preceding the birth of game theory @xcite , in which two players can virtually cooperate and get the optimal profit at the maximal quantum entanglement between them albeit both act `` selfishly '' just as they do in the classical case .
this work actually set up a foundation for further discussions regarding quantum game with continuous strategies @xcite .
recently qin _ et al . _
introduced a quantization scheme for asymmetric games @xcite .
the key modification of their scheme compared with li _
s is the adoption of two different entanglement factors , which ensure the asymmetric games to get the optimal cooperative profit .
nevertheless , qin _
s quantum scheme can still be regarded as a fair play , for the two factors are determined by the extent of the asymmetry of the game .
it is quite often that various environments may affect the quantum game , leading to asymmetric forms of the entanglement operator . hence , to better understand the property of quantum game , it s worthwhile studying the distinctive features of quantum game resulted from the change of the entanglement operator . in this letter
we develop the quantum model of cournot s duopoly proposed by li _
by changing the initial state before actually entangling them , and we define these operations altogether as a whole entangling operation .
thus it can be discovered that the total profit at the unique nash equilibrium will increase monotonously as long as the entropy of the entangled state , namely the entanglement , tends to infinity .
while , considering the lack of symmetry in this quantization scheme , our result will not remain pareto optimal ( the best result that can be achieved without disadvantaging at least one group ) , which in fact may not be a necessity in the real world because of the simple fact that there is no absolutely - impartial law . besides , all of the operators we cite in the quantization of cournot duopoly can be realized experimentally within the capacity of modern optical technology through proper design .
now we briefly recall the classical cournot duopoly .
duopoly is an economic condition in which two firms hold the market of a certain commodity without a third competitor . in cournot model ,
two firms are assumed to produce a homogeneous product and simultaneously decide their quantities @xmath0 and @xmath1 . and the price of this product is decided by the quantities as @xmath2 where @xmath3 is the total quantity of two firms , i.e. @xmath4 and @xmath5 is a constant .
assume that the unit cost of the product is also a constant @xmath6 , the profit function of each firm would be : @xmath7 with @xmath8 and @xmath9 .
it s easy to work out the unique nash equilibrium : @xmath10 however , the nash equilibrium is not the optimal solution for the market . if the two firms cooperate and raise their total quantity to @xmath11 , they can acquire _ the highest total profit _ @xmath12 .
in addition , by keeping the symmetry of the model , we can obtain the pareto optimal in the following solution : @xmath13 the profit difference between the optimal choice and the nash equilibrium reflects the conflict between individual rationality and collective rationality , just like in the classical prisoners dilemma .
here we put forward a modified quantization model of li _ et al . _
s @xcite , where we make use of two single - mode electromagnetic fields .
[ structure ] shows the quantum structure of the game . as a extension ,
we first apply two single - mode squeezing operator @xmath14 to the vacuum state , where @xmath15 , and @xmath16 are the creation and annihilation operator of the corresponding field , followed by a two - mode squeezing operator @xmath17 of a two - mode electromagnetic field .
thus the total entanglement operator would be : @xmath18 \notag \\ & \cdot \exp [ \gamma _ { 1}(a_{1}{}^{2}-a_{1}^{\dag } { } ^{2})/2 ] \notag \\ & \cdot \exp [ \gamma _ { 2}(a_{2}{}^{2}-a_{2}^{\dag } { } ^{2})/2],\end{aligned}\ ] ] with constraint @xmath19 and @xmath20 ( which also is an extension of li _ et al _ s model ) .
apparently , we can turn back to li _
et al _ s entanglement operator if @xmath21 . the strategies in the set of local unitary operators is @xmath22 $ ] as the scheme of li _ et al .
the final measurement is made corresponding to the observables @xmath23 , which in quantum optics is the amplitude quadrature ( we can simply perceive it as the `` position '' operator ) .
this measurement is done by the homodyne measurement assuming that the state is infinitely squeezed .
correspondingly , by citing the phase quadrature @xmath24 ( which can be seen as the `` momentum '' operator ) , we can rewrite the entanglement operator and the strategies in the following form : @xmath25 \notag \\ & \quad \,\cdot \exp [ i\gamma _ { 1}(\hat{x}_{1}\hat{p}_{1}+\hat{p}_{1}\hat{x}% _ { 1})/2 ] \notag \\ & \quad \,\cdot \exp [ i\gamma _ { 2}(\hat{x}_{2}\hat{p}_{2}+\hat{p}_{2}\hat{x}% _ { 2})/2 ] , \\ \hat{d}_{j}(x_{j } ) & = \exp ( -ix_{j}\hat{p}_{j}),\qquad j=1,2.\end{aligned}\ ] ] obviously , the classical cournot duopoly is only a special case with @xmath26 in its quantum extension .
because the strategy set is not extended compared to its classical counterpart , we can be sure that all the features shown in the quantization version attribute to the entanglement of the states . _ the specific property of the entangled state is analyzed in the paragraphs below . _
by detailed calculation we get : @xmath27 , \\
\hat{j}^{\dag } \hat{d}_{2}(x_{2})\hat{j } & = \exp [ -ix_{2}(\hat{p } _ { 2}e^{\gamma _ { 2}}\cosh \gamma _ { 12}+\hat{p}_{1}e^{\gamma _ { 1}}\sinh \gamma _ { 12})].\end{aligned}\ ] ] therefore the final state reads : @xmath28\left\vert vac\right\rangle _ { 1 } \notag \\ & \otimes \exp [ -ie^{\gamma _ { 2}}(x_{2}\cosh \gamma _ { 12}+x_{1}\sinh \gamma _ { 12})\hat{p}_{2}]\left\vert vac\right\rangle _ { 2}.\end{aligned}\ ] ] and the measurement gives the respective quantity of each firm player : @xmath29 simply by substituting the above into the classical profit functions ( see eq .
( cl - profit ) ) , we get the quantum profit : @xmath30 where @xmath31 $ ] . solving the nash equilibrium equations we obtain the unique solution :
@xmath32 and the profit at the equilibrium reads : @xmath33 where @xmath34 .
apparently , the profit functions only depend on two parameters @xmath35 and @xmath36 .
in addition , if we make @xmath37 , i.e. @xmath38 , it s easy to find that the result is exactly that of li _ et al _ in their quantization scheme .
using the method put forward in rendell s paper @xcite , we write the entangled state in the representation of `` position '' : @xmath39.\ ] ] by solving the equations below : @xmath40 we can obtain the expression of the three unknown parameters as @xmath41 where @xmath42 . to better assess the entanglement of the state
, we need to calculate the von neumann entropy of the state , which is defined as @xmath43tr@xmath44tr@xmath45 ( @xmath46 are the density matrix of two single - mode electromagnetic fields ) . based on the above calculation we get the following equality by making use of rendell s method : @xmath47 where @xmath48 . as a confirmation , when @xmath37 , eq .
( [ entropy ] ) is reduced to : @xmath49 which is exactly the entropy of a two - mode squeezing state .
it is interesting to find that like the profit function , the entropy is also determined by only two parameters @xmath35 and @xmath50 . in addition , we find that the von neumann entropy monotonously increases from zero to infinity as the parameter @xmath51 increases . since @xmath51 increases as either of the two parameters increases ( when @xmath52 ) , the entropy is an increasing function of both @xmath35 and @xmath36 . in li _
et al . _ s entanglement operator , through _ whole unitary transformation _ @xmath53 $ ] , we get the maximal entangled state as @xmath35 tends to infinity . nevertheless , it s discovered that only by changing the initial state via _ local unitary transformation _
$ ] , the maximized entanglement is also achieved as long as entanglement exists ( i.e. @xmath52 , even if it is very small ; see fig .
[ r12-fixed ] ) .
obviously , such kind of situation is impossible in quantum game with hilbert space of finite dimensions .
moreover , we estimate the asymmetry by calculating the relative difference of the `` position '' uncertainties ( which actually represent the width of gaussian wave package in each `` position '' space ) between two parties in the expression below : @xmath55 for fixed and finite @xmath35 , the above expression would monotonously decreases as @xmath36 increases . due to the obvious fact that the system is completely symmetric when @xmath38 , we can simply perceive @xmath56 as a measurement of the asymmetry of the entangled state , while it s worth mentioning that all the discussion concerning asymmetry above only reflects a variation trend between variables and the final state .
so far , we have confined three parameters presented in the entanglement operator to only two ones , i.e. @xmath57 . in the paragraphs below ,
specific analysis of the relation between the game model adopted here and the entangled state is given .
since the total profit of two players can not achieve the maximum at the nash equilibrium due to the deviation between individual rational choice and collective optimization requirement in classical cournot duopoly , it must be the application of quantum entanglement that eliminates this dilemma in li _
et al . _ s scheme . as an extension of their entangled state
, we separately give the alternation trends of the total profit ( @xmath58 ) and the entropy @xmath59 versus @xmath60 in the following figures ( fig .
[ dr - fixed ] , fig .
[ r12-fixed ] ) , in which the unit of profit is taken as @xmath12 , namely , the optimal value of the total profit in the classical cooperation scenario ( see eq .
( [ cl - co - result ] ) ) .
it s obvious to see that both the profit and entropy function increases as either of the parameters increases ( the condition @xmath52 is necessary here ; otherwise there would be no entanglement between two parties ) .
when the entropy tends to infinity , the total profit also achieves its maximal value , and li _ et al .
_ s situation is just a special case when @xmath38 .
since the entropy is defined as a measurement of the entanglement of the state , it would be adequate to explain that the entanglement of the state does help improve the total profit in cournot duopoly independent of the concrete form of the state ( i.e. no matter in which way the entanglement is increasing ) as illustrated in the previous section , and so long as _ the maximal entanglement _
is attained , the conflict between individual rationality and collective rationality disappears . in fig .
[ s - u ] , we present a more detailed function plot between the entropy and the total profit . apparently , although both quantities have the same increasing trend as seen in fig .
[ dr - fixed ] and fig .
[ r12-fixed ] , the entropy does not determine the total profit uniquely , i.e. there is nt a one - to - one correspondence between such two quantities .
in combination of previous discussion concerning asymmetry , it is not difficult to find that _ the increase of asymmetry _ would cause the reducing of the total profit if the entanglement of the state do nt reach the maximum .
furthermore , the variation of asymmetry also results in the change of the profit difference between two players in our quantization scheme .
[ effect - of - dr ] shows that player 1 would monopolize the market and get the optimal profit when @xmath61 tends to positive infinity , while on the contrary , player 2 would become the monopolizer of the market if @xmath50 tends to negative infinity .
therefore , @xmath62 can be separately regarded as the extent to which player 1 and 2 are encouraged ( or the corresponding opponents are suppressed ) , that is , the larger the parameter is , the more profit the corresponding player can acquire .
we investigate the quantization of games with continuum strategic space by making use of a more general entanglement operator . for the particular case of symmetric scenario
, it would turn back to the quantization scheme put forward by li _
et al._. as an extension of their quantum model , we investigate _ the relation between the nash equilibrium in quantum situation and the entanglement of the state _ more specifically , and find that the optimal total profit can always be attained as long as the maximal entanglement is realized .
also we observed some novel features such as the encouraging and suppressing effect in our quantum model , which completely attributes to _ the asymmetry of the system_. therefore , in our quantization scheme , the government would have more choices in managing the market .
it actually constitutes a better base for further discussion of game models involving incomplete information or unequal competition status .
this work is supported by the national fundamental research program ( grant no .
2001cb309300 ) , the national science fund for distinguished young scholars ( grant no .
10425524 ) and the project in quantum information technology ( grant no .
r-144 - 000 - 071 - 305 ) . | in the present letter , we propose a more general entangling operator to the quantization of cournot economic model , in which players can access to a continuous set of strategies . by analyzing the relation between the von neumann entropy of the entangled state and the total profit of two players precisely , we find that the total profit at the nash equilibrium always achieves its maximal value as long as the entropy tends to infinity . moreover , since the asymmetry is introduced in the entangled state , the quantum model shows some kind of `` encouraging '' and `` suppressing '' effect in profit functions of different players . |
for more than two decades , the optical process of parametric down - conversion ( pdc)@xcite has been the primary photon source in quantum optics experiments@xcite .
this nonlinear process converts single photons into pairs of signal and idler photons , which possess quantum correlations between their various degrees of freedom .
the statistical nature of this process is a subject of great interest , as the photon - number statistics of the process provide a direct indication of its non - classility@xcite .
additionally , the photon statistics have practical implications on state preparation schemes , where the quality of the produced state depends on the number of photons generated simultaneously within the pdc process@xcite .
recent developments in photon - number resolving detectors@xcite enable the implementation of a wide range of applications in quantum optics , including state preparation schemes@xcite , quantum state filtering@xcite and quantum gates@xcite .
number discriminating capabilities have also been shown to aid in probing eavesdropping attacks on quantum cryptography protocols@xcite .
furthermore , one of the most prominent advantages of number - resolving detectors is the ability to directly measure the photon - number statistics of an optical state .
the photon - number distribution of a pdc source can range between a thermal distribution , when a single down - converted mode is collected , and a poisson distribution when an infinitely large number of spatial and spectral modes ( multimode ) is collected@xcite .
the photon statistics of pdc sources has been directly measured in several works@xcite , all of which report a poisson distribution .
however , the majority of quantum optics schemes take great care to ascertain that only a single mode is collected . using an indirect homodyne detection method ,
a thermal distribution from a parametric amplifier was demonstrated@xcite .
transition between thermal to poisson distributions was indirectly observed through the @xmath0 parameter@xcite and through high - order pdc events@xcite . in this work
, we use a photon - number resolving detector in order to directly measure the photon statistics of the signal and idler photons collected from a controlled number of modes , allowing us to observe the range of distributions between the single - mode and the multimode extremes .
the photon - number resolving detector used in this work is the silicon photomultiplier ( sipm ) @xcite .
the sipm detector is composed of multiple avalanche - photo - diodes ( apd ) operating in geiger - mode , combined on a single substrate .
when a photon impinges on an apd element , an electric discharge can be generated .
the photons impinging on the detector are distributed across its many elements .
the output signals generated by all apds are combined to form a single output pulse whose intensity is proportional to the number of triggered elements . in comparison to other technologies ,
the sipm detector offers a better photon - number resolution ( up to 20 photons , compared to 10 or less ) , easy integration in optical setups and operation at room temperature . on the other hand ,
its photon detection efficiency is 26 times lower , depending on the working wavelength .
the photon statistics measured by sipm detectors may deviate from that of the original state .
the three main reasons for these deviations are : 1 ) the non - perfect overall photon detection efficiency ( loss ) , which is determined by the detection and coupling efficiencies , 2 ) false detections caused by thermal excitations ( dark counts ) , and 3 ) optical crosstalk , in which secondary photons , created by carrier relaxation in one apd element , are detected by a neighboring element , falsely increasing the photon count@xcite .
another deviation is caused by events when two or more photons impinge on the same element and are detected as one .
however , we have shown using a numerical model that this effect becomes significant only when the number of detected photons is higher than 20% of the total number of elements@xcite .
the descriptions of loss and dark counts are quite straightforward .
however , several currently available models for the crosstalk effect are either computationally difficult@xcite or only applicable under limiting conditions@xcite .
we introduce here an analytical crosstalk model which can be applied for the entire range of experimental settings .
this paper is organized as follows : in sec .
[ sec : data - analysis ] we present a model for the distortion effects in the sipm detector , that includes a new approach to the crosstalk effect .
this model is used for the interpretation of experimental data and the reconstruction of the original photon - number statistics .
[ sec : experimental ] describes the experimental setup for the generation of pdc states and their detection using an sipm photon - number resolving detector . in sec .
[ sec : results-&-discussion ] we present measurements of photon - number distributions of parametric down - conversion and their dependence on the number of collected modes .
conclusions are presented in sec .
[ sec : conclusions ] .
in order to interpret the photon - number statistics measured using sipm detectors , the effects of loss , dark counts , and optical crosstalk must be properly modeled . if the measured photon - number distribution is represented by a vector of probabilities @xmath1 , the original probability distribution @xmath2 , is related to the measured distribution through the relation @xmath3 each component of the distortion matrix @xmath4
is the conditional probability @xmath5 of measuring @xmath6 photons given that @xmath7 original photons arrived at the detector . using eq .
[ eq : measured - orig relation ] , the original distribution can be reconstructed , in principle , by multiplying the measured distribution with the inverse of the distortion matrix .
this method , however , is limited to relatively large probabilities and may produce non - physical results otherwise@xcite .
another approach , which we use here , is to apply a fitting algorithm which finds the original distribution based on a goodness of fit test .
the matrix @xmath4 can be written as a product of three matrices , @xmath8 , which represent the loss , dark counts ( dk ) and crosstalk ( _ ct _ ) effects .
the individual matrices are constructed as follows : the probability for loss is described by the relation@xcite @xmath9 where @xmath10 is the overall detection efficiency .
the probability of dark counts is assumed to follow a poisson distribution@xcite of the form @xmath11 where @xmath12 is the average number of dark counts .
optical crosstalk occurs when a secondary photon generated during an electronic discharge in one apd element is detected in a neighboring apd element , thereby creating a secondary avalanche .
photons produced in the secondary avalanches can then continue to trigger further avalanches in additional elements ( crosstalk - generated - crosstalk ) .
the crosstalk probabilities in our model are calculated by dividing the crosstalk events into stages and recursively counting the number of possible crosstalk events in each stage .
the recursion formula for the crosstalk matrix element is : @xmath13 where @xmath14 is defined as the overall probability for a crosstalk - avalanche to be generated in any of the four neighboring elements .
each stage has @xmath7 triggered elements which can trigger up to @xmath7 new crosstalk events .
the probability of generating @xmath15 new crosstalk events is composed of the probability that @xmath15 elements out of the @xmath7 possible will generate crosstalk , while the remaining @xmath16 elements will not .
this probability also includes the combinatorial factor of choosing the @xmath15 crosstalk - generating elements out of the possible @xmath7 .
the @xmath15 crosstalk - triggered elements continue to generate additional crosstalk events in a recursive way , until all of the @xmath17 crosstalk elements are triggered .
this novel analytical crosstalk model does not limit the number of crosstalk events and also accounts for crosstalk - generated - crosstalk .
the model was developed since other available analytical models @xcite were inconsistent with our experimental results due to limitations on the crosstalk probability values and the number of crosstalk events .
our model does not account for the geometrical arrangement of the apd elements in the detector , nor its finite size .
however , we show below that for values within the experimental range , this analytical model agrees with a numerical model that has no such limitations@xcite .
additionally , this model is less computationally demanding than the numerical model .
thus , it is a useful tool in the interpretation of experimental data from photon - number resolving detectors , where crosstalk between the detection elements plays a crucial role .
we have measured the photon - number distributions of the polarization modes of a stimulated type - ii collinear pdc process . the experimental setup is shown in fig .
[ fig : setup](a ) .
pulses of 150fs at a repetition rate of 250khz are amplified , frequency doubled to 390 nm and used to pump a 2 mm long type - ii collinear @xmath18 ( bbo ) nonlinear crystal .
the degenerate wavelength signal and idler photons at 780 nm , which are created with orthogonal polarizations ( horizontal ( h ) and vertical ( v ) ) , are split at a polarizing beam splitter ( pbs ) and detected separately using two photon - number resolving detectors .
this work focuses on measuring the photon - number distribution of a single polarization mode .
thus , the detection of both polarization modes is only used to evaluate the heralded detection efficiency of the sipm detector .
the results presented for a single polarization mode also apply to the photon pair statistics , as a collinear process produces the same statistics for pairs and for individual photons . before the photons are coupled to the detectors , they are spectrally and spatially filtered by bandpass filters and by optical fibers , respectively . the number of spatial modes which are collected can be changed by using optical fibers with different core diameters and numerical apertures@xcite , and the number of collected spectral modes can be changed by using filters of different bandwidths . the detection configuration is shown in fig . [
fig : setup](b ) . the optical signal from the photon source
is coupled to an sipm detector with 100 elements ( _ hamamatsu photonics _ , s10362 - 11 - 100u ) using optical fibers .
the output electrical signal from the detector is amplified using low - noise amplifiers , digitized and then analyzed by fpga electronics , where the number of detected photons is extracted in real time .
this data is continuously transmitted to a computer , which presents the statistics . in order to minimize the effects of dark counts and after - pulsing@xcite
, the 1 ns long sampling time of the analog electrical signal is synchronized with the arrival time of the photons .
the dark counts are further reduced by moderately cooling the detectors using a thermoelectric cooler to @xmath19 , and the bias voltage level applied to the sipm is modified accordingly . ]
figure [ fig : histogram ] shows an example of a histogram of pulse intensities produced by a coherent state .
the probability of detecting @xmath6 photons is proportional to the area of the @xmath20 gaussian .
we obtain high photon - number discrimination for as high as 20 photons .
the photon - number resolution error , which results from the overlapping between neighboring gaussians , is smaller than 1% below 12 photons and approaches 12% for 20 photons .
the number of resolvable photons is limited due to the decrease of the signal to noise ratio as the number of photons is increased . ]
we demonstrate the use of our model for reconstructing the original photon statistics with the photon - number distribution measurements of a thermal state .
the thermal state was generated by collecting a single polarization mode of pdc using the setup shown in fig .
[ fig : setup ] . the original thermal distribution can be written as @xmath21 where @xmath22 is the average number of photons .
the measured distribution exhibits a lower value of @xmath22 due to losses and deviates from the power - law dependence of eq .
[ eq : pth_thermal ] due to optical crosstalk and dark counts .
we measured the same thermal state with three different sets of values for the amount of loss , dark counts and crosstalk , obtained by operating the sipm detector with different bias voltage values@xcite .
as a result , we obtained three different measured distributions , although the original photon - number distribution was the same .
the measurements are shown in fig .
[ fig : model fits ] .
the main distortions in the distributions are the low values of @xmath22 due to losses and the deviation from a straight line in the semi - log plot when the number of measured photons exceeds 1 .
the latter is the effect of the crosstalk process , which falsely increases the probability of measuring high numbers of photons .
the crosstalk probabilities of these measurements range from the highest probability value ( maximum bias voltage ) , which creates the highest deviation , to the minimal probability ( minimal bias voltage ) which creates the lowest deviation . .
the bias voltage values and the fit parameters are : @xmath23 , @xmath24 , @xmath25 , @xmath26 , @xmath27 ( green circles ) , @xmath28,@xmath29 , @xmath30 , @xmath31 , @xmath32 ( red upward triangles ) , and @xmath33 , @xmath34 , @xmath35 , @xmath36 , @xmath37 ( blue downward triangles ) . the corresponding crosstalk values obtained with the numerical model of ref.@xcite are @xmath38 , @xmath39 , and @xmath40 , respectively . fits using the numerical model are presented by black dashed lines.[fig : model fits ] ] the original distribution is reconstructed by fitting the measured data @xmath1 to the function @xmath41 , where @xmath42 is the thermal distribution of eq .
[ eq : pth_thermal ] and @xmath22 is the only free parameter .
the overall detection efficiency @xmath10 , the average number of dark counts @xmath12 , and the crosstalk probability @xmath14 , that are used for the construction of the distortion matrices , are measured separately .
the value of @xmath10 is defined as the heralded efficiency between the two down - converted polarization modes and is determined by the ratio between two - photon coincidence and single - photon counts in the limit @xmath43 .
the dark counts and crosstalk parameters are obtained through a separate measurement of the poissonian dark count statistics which are fit to the function @xmath44 , where @xmath45 is a vector of zeros with 1 at the first position , and @xmath12 and @xmath14 are the free fit parameters . all three measurements in fig . [ fig : model fits ] were reconstructed to the same original state , within the margins of error , despite the different measured distributions .
this result indicates the consistency of our evaluation of the dark counts , loss and crosstalk parameters .
we compare these results to those obtained using the numerical crosstalk model introduced in ref.@xcite .
the numerical model is based on simulations of the detection process , and takes into account the geometrical configuration of the detector , the finite number of detection elements , and possible attempts to trigger neighboring cells which have already been triggered .
the model is characterized through a parameter @xmath46 , defined as the probability of generating crosstalk in one particular neighbor among the four nearest neighbors , rather than the overall probability of generating crosstalk among the nearest neighbors , which we define here as @xmath14 .
the values of @xmath46 obtained from the fit correspond to the values of @xmath14 through the relation @xmath47@xcite .
the agreement between the two models shows that although the analytical model does not account for the geometry of the detector , it provides a good description of the crosstalk process for all crosstalk values in the experimental range .
we now consider the photon - number distributions of down - converted photons as a function of the number of collected modes . if @xmath6 photons are distributed among @xmath48 modes , such that @xmath49 , then the overall probability of measuring @xmath6 photons is given by@xcite : @xmath50 where the product @xmath51 is the joint probability of measuring @xmath52 photons in each mode @xmath53 , and the combinatorial factor amounts to the number of possible arrangements of @xmath6 indistinguishable photons into @xmath48 modes . if the photons are distributed evenly among the modes , the average number of photons in each mode is uniform and equals @xmath54 the photon - number distribution within each of the modes is thermal , @xmath55 , and by substituting eq .
[ eq : pth_thermal ] into eq .
[ eq : ps(n ) ] , we obtain the following probability distribution @xmath56 known as the negative binomial distribution .
this distribution converges to a poisson distribution when the number of modes approaches infinity .
, @xmath57 , @xmath58 , 3 nm bandwidth filter , single mode fiber for 780 nm , integration time of 780minutes ( green circles ) , @xmath59 , @xmath60 , @xmath61 , 3 nm bandwidth filter , single mode fiber for 1500 nm , integration time of 100minutes ( blue squares ) , @xmath62 , @xmath63 , @xmath64 , 10 nm bandwidth filter , single mode fiber for 1500 nm , integration time of 100minutes ( orange triangles ) , and @xmath65 , @xmath66 , @xmath67 , 3 nm bandwidth filter , multimode fiber for 780 nm , integration time of 115minutes ( red diamonds ) .
[ fig : mode graph ] ] the dependence of the photon - number distribution on the number of modes is demonstrated by measuring the photon statistics collected from a controlled number of modes .
figure [ fig : mode graph ] shows four representative measurements of the photon - number distributions which range between a single mode ( thermal ) distribution , and a multimode ( poisson ) distribution .
the solid lines are fits to the function @xmath68 , where @xmath69 is defined in eq .
[ eq : ps(n ) thermal ] with @xmath22 and @xmath48 as the free parameters .
we detected up to 14 photons and obtained good fits to the data .
notice that the difference between the distributions becomes more marked as the number of photons increases ( see figs .
[ fig : mode graph](b)(e ) ) .
the difference is almost undetectable for low photon numbers .
in fact , measurements of at least 10 photons are required for our experimental parameter range , in order to properly discriminate between distributions with different mode numbers .
the number of modes obtained for all combinations of fibers and filters is shown in table [ tab : mode table ] .
the non - integer mode numbers should be interpreted as a weighted number of occupied modes and result from a non - uniform distribution of photons between the modes .
we can discriminate between distributions containing up to about 10 modes , as distributions containing over 10 modes can not be distinguished from poisson distributions .
table [ tab : mode table ] shows the proportional increase of the number of spectral modes with the increase in the bandwidth of the bandpass filters .
the width of the phase - matched spectrum of the down - converted photons is estimated to be @xmath70 . due to spectral distinguishability
, this width includes about 3 temporal modes@xcite .
the number of spatial modes measured with our system is also consistent with the choice of fibers .
when the pdc photons , which passed through a 3 nm bandpass filter , were coupled into a single mode fiber for 780 nm , a thermal distribution was produced . whereas , when the photons were coupled into a multimode fiber , a poisson distribution was produced , even though the same filter was used .
the relatively small number of modes obtained with the single - mode fiber for 1550 nm can be attributed to non - uniform mode coupling , which does not equally stimulate all of the modes in the fiber .
5in@>p0.6in > p0.6in|>p0.7in
> p0.7in > p0.7in >
p0.7 in & + mfd & na & 3 nm & 5 nm & 10 nm & no filter + @xmath71 & 0.13 & @xmath72 & @xmath73 & @xmath74 & @xmath75 + @xmath76 & 0.14 & @xmath77 & @xmath78 & @xmath79 & @xmath80 + @xmath81 & 0.275 & poissonian & & & + the stimulated nature of pdc can also be used to obtain the number of collected modes@xcite .
the average number of down - converted photons in a single mode is determined by the stimulation parameter @xmath82 , which has a linear dependence on the pump field , the crystal length and the nonlinear coefficient of the crystal@xcite .
assuming an equal distribution of the photons between @xmath48 modes , the mean number of photons @xmath22 is given by @xmath83 where @xmath84 is the average number of photons in each of the individual modes , @xmath85 is the coupling constant between the pump field and the nonlinear crystal , and @xmath86 is the pump intensity . in order to show the dependence of @xmath22 on the number of modes , we measured photon - number distributions as a function of the pump intensity for the different numbers of collected modes .
the values of @xmath22 were extracted by fitting the different distributions to eq .
[ eq : ps(n ) thermal ] .
the fit values of @xmath22 as a function of the total pump power are shown in fig .
[ fig : stimulation ] .
all measurements show a nonlinear dependence , indicative of a stimulated process .
the number of modes @xmath48 obtained from these fits corresponds well to the number of modes obtained through fits to the photon - number distributions ( see fig . [
fig : mode graph ] ) .
furthermore , all fits resulted in similar values of the @xmath85 parameter , in agreement with the model of eq .
[ eq : ns ] . with the following parameters : ( a ) @xmath87 , @xmath88 ( green circles ) ( b ) @xmath89 , @xmath90 ( blue squares ) ( c ) @xmath91 , @xmath92 ( orange triangles ) .
( d ) @xmath93 , @xmath94 ( red diamonds).[fig : stimulation ] ]
in conclusion , we have built a detection setup , which incorporates a silicon photomultiplier ( sipm ) as a photon - number resolving detector .
our setup enables good photon - number discrimination for as many as 20 photons .
the output signal of the sipm is analyzed in real time using fpga electronics .
the measured photon - number distribution probabilities are interpreted by modeling the effects of loss , dark counts and optical crosstalk , which distort the original photon - number probabilities .
we present a useful analytical crosstalk model that is applicable for all crosstalk probability values in the experimental range and includes crosstalk - generated - crosstalk .
we have performed measurements of photon - number distributions of a single polarization mode of a collinear type - ii pdc process .
the high photon - number resolution of our system enables the differentiation between distributions collected from a different number of spectral and spatial modes .
we present distributions which range from single - mode ( thermal ) to multimode ( poissonian ) statistics .
the results show how the photon - number distribution is determined by the number of collected modes .
furthermore , we show the dependence of the average number of photons on the number of collected modes , and the stimulated nature of parametric down - conversion . | optical parametric down - conversion ( pdc ) is a central tool in quantum optics experiments .
the number of collected down - converted modes greatly affects the quality of the produced photon state .
we use silicon photomultiplier ( sipm ) number - resolving detectors in order to observe the photon - number distribution of a pdc source , and show its dependence on the number of collected modes .
additionally , we show how the stimulated emission of photons and the partition of photons into several modes determine the overall photon number .
we present a novel analytical model for the optical crosstalk effect in sipm detectors , and use it to analyze the results .
10 l. mandel and e. wolf , _ optical coherence and quantum optics _ ( cambridge university press , 1995 ) .
p. g. kwiat , k. mattle , h. weinfurter , a. zeilinger , a. v. sergienko , and y. shih , `` new high - intensity source of polarization - entangled photon pairs , '' phys .
rev .
lett . * 75 * , 43374341 ( 1995 ) .
e. waks , e. diamanti , b. c. sanders , s. d. bartlett , and y. yamamoto , `` direct observation of nonclassical photon statistics in parametric down - conversion , '' phys .
rev .
lett . * 92 * , 113602 ( 2004 ) .
e. waks , b. c. sanders , e. diamanti , and y. yamamoto , `` highly nonclassical photon statistics in parametric down - conversion , '' phys .
rev .
a * 73 * , 033814 ( 2006 ) .
o. a. ivanova , t. s. iskhakov , a. n. penin , and m. v. chekhova , `` multiphoton correlations in parametric down - conversion and their measurement in the pulsed regime , '' quantum electronics * 36 * , 951 ( 2006 ) .
c. t. lee , `` nonclassical photon statistics of two - mode squeezed states , '' phys .
rev .
a * 42 * , 16081616 ( 1990 ) .
h. s. eisenberg , g. khoury , g. a. durkin , c. simon , and d. bouwmeester , `` quantum entanglement of a large number of photons , '' phys .
rev .
lett . * 93 * , 193901 ( 2004 ) .
d. lincoln , `` a large statistics study of the performance and yields of generation-6 vlpcs ( histe - vi ) , '' nuclear instruments and methods in physics research section a : accelerators , spectrometers , detectors and associated equipment * 453 * , 177 181 ( 2000 ) .
d. achilles , c. silberhorn , c. sliwa , k. banaszek , i. a. walmsley , m. j. fitch , b. c. jacobs , t. b. pittman , and j. d. franson , `` photon - number - resolving detection using time - multiplexing , '' journal of modern optics * 51 * , 14991515 ( 2004 ) .
d. rosenberg , a. e. lita , a. j. miller , and s. w. nam , `` noise - free high - efficiency photon - number - resolving detectors , '' phys .
rev .
a * 71 * , 061803 ( 2005 ) .
b. e. kardynal , z. yuan , and a. j. shields , `` an avalanche photodiode - based photon - number - resolving detector , '' nat .
photonics * 2 * , 425 ( 2008 ) .
e. dauler , a. kerman , b. robinson , j. yang , b. voronov , g. goltsman , s. a. hamiltom , and k. berggren , `` photon - number - resolution with sub-30-ps timing using multi - element superconducting nanowire single photon detectors , '' j. mod . opt . * 56 * , 365 ( 2009 ) .
p. kok , h. lee , and j. p. dowling , `` creation of large - photon - number path entanglement conditioned on photodetection , '' phys .
rev .
a * 65 * , 052104 ( 2002 ) .
y. gao , p. m. anisimov , c. f. wildfeuer , j. luine , h. lee , and j. p. dowling , `` super - resolution at the shot - noise limit with coherent states and photon - number - resolving detectors , '' j. opt .
soc . am .
b * 27 * , a170a174 ( 2010 ) .
h. s. eisenberg , j. f. hodelin , g. khoury , and d. bouwmeester , `` multiphoton path entanglement by nonlocal bunching , '' phys .
rev .
lett .
* 94 * , 090502 ( 2005 ) .
r. okamoto , j. l. obrien , h. f. hofmann , t. nagata , k. sasaki , and s. takeuchi , `` an entanglement filter , '' science * 323 * , 483485 ( 2009 ) .
e. knill , r. laflamme , and g. milburn , `` a scheme for efficient quantum computation with linear optics , '' nature * 409 * , 46 ( 2001 ) .
t. b. pittman , b. c. jacobs , and j. d. franson , `` demonstration of nondeterministic quantum logic operations using linear optical elements , '' phys .
rev .
lett . * 88 * , 257902 ( 2002 ) .
g. a. durkin , c. simon , and d. bouwmeester , `` multiphoton entanglement concentration and quantum cryptography , '' phys .
rev . lett . * 88 * , 187902 ( 2002 ) .
l. mandel , `` fluctuations of photon beams : the distribution of the photo - electrons , '' proceedings of the physical society * 74 * , 233 ( 1959 ) .
f. paleari , a. andreoni , g. zambra , and m. bondani , `` thermal photon statistics in spontaneous parametric downconversion , '' opt . express * 12 * , 28162824 ( 2004 ) .
m. avenhaus , h. b. coldenstrodt - ronge , k. laiho , w. mauerer , i. a. walmsley , and c. silberhorn , `` photon number statistics of multimode parametric down - conversion , '' phys
. rev .
lett .
* 101 * , 053601 ( 2008 ) .
m. vasilyev , s .- k .
choi , p. kumar , and g. m. dariano , `` investigation of the photon statistics of parametric fluorescence in a traveling - wave parametric amplifier by means of self - homodyne tomography , '' opt
. lett .
* 23 * , 13931395 ( 1998 ) .
a. eckstein , a. christ , p. j. mosley , and c. silberhorn , `` highly efficient single - pass source of pulsed single - mode twin beams of light , '' phys .
rev .
lett . * 106 * , 013603 ( 2011 ) .
w. wasilewski , c. radzewicz , r. frankowski , and k. banaszek , `` statistics of multiphoton events in spontaneous parametric down - conversion , '' phys .
rev .
a * 78 * , 033831 ( 2008 ) .
g. bondarenko , b. dolgoshein , v. golovin , a. ilyin , r. klanner , and e. popova , `` limited geiger - mode silicon photodiode with very high gain , '' nucl .
phys . b. ( proc .
suppl ) * 61 * , 347 ( 1998 ) .
p. buzhan , b. dolgoshein , l. filatov , a. ilyin , v. kaplin , a. karakash , s. klemin , r. mirzoyan , a. otte , e. popova , v. sosnovtsev , and m. teshima , `` large area silicon photonmultipliers : performance and appli , '' nucl .
instr . and meth . a * 567 * , 78 ( 2006 ) .
l. dovrat , m. bakstein , d. istrati , and h. eisenberg , `` simulations of the detection process in silicon photomultiplier detectors , '' arxiv:1109.0698v1 ( 2011 ) .
i. afek , a. natan , o. ambar , and y. silberberg , `` quantum state measurements using multipixel photon detectors , '' phys .
rev .
a * 79 * , 043830 ( 2009 ) .
m. akiba , k. tsujino , k. sato , and m. sasaki , `` multipixel silicon avalanche photodiode with ultralow dark count rate at liquid nitrogen temperature , '' opt .
express * 17 * , 1688516897 ( 2009 ) .
p.
eraerds , m. legr , a. rochas , h. zbinden , and n. gisin , `` sipm for fast photon - counting and multiphoton detection , '' opt .
express * 15 * , 1453914549 ( 2007 ) .
h. lee , u. yurtsever , p. kok , g. m. hockney , c. adami , s. l. braunstein , and j. p. dowling , `` towards photostatistics from photon - number discriminating detectors , '' journal of modern optics * 51 * , 15171528 ( 2004 ) .
a. yariv , _ optical electronics in modern communications _ ( oxford university press , 1991 ) . p. j. mosley , j. s. lundeen , b. j. smith , p. wasylczyk , a. b. uren , c. silberhorn , and i. a. walmsley , `` heralded generation of ultrafast single photons in pure quantum states , '' phys .
rev .
lett .
* 100 * , 133601 ( 2008 ) .
p.
kok and s. l. braunstein , `` postselected versus nonpostselected quantum teleportation using parametric down - conversion , '' phys
. rev .
a * 61 * , 042304 ( 2000 ) . |
the low temperature relaxation of the magnetization of magnetic molecular solids such as 8 has proven difficult to understand ever since the earliest experimental studies @xcite .
the time dependence of this relaxation is highly non - exponential , and fits to forms such as stretched exponentials have provided no insight even when the fits seem to be good .
a second puzzling feature is that for short times , the relaxation is observed to follow a square - root behavior with time in a large number of protocols : demagnetization , magnetization , and hole - digging in which the magnetic field is abruptly changed after the magnetization has been allowed to come to an equilibrium or quasi - equilibrium state in response to a previous value of the applied magnetic field .
a good review of the subject is given by gatteschi , sessoli , and villain @xcite .
these authors give many more references to experimental studies @xcite , theoretical analyses @xcite , and monte carlo simulations @xcite .
the fundamental microscopic mechanism by which the spin of an individual molecule changes at low temperatures ( say below 50 mk ) is incoherent tunneling between the lowest energy states . in both 8 and 12 ,
the anisotropy of the molecule is of the ising type , and the lowest energy states have zeeman quantum numbers @xmath0 , where @xmath1 is the spin of the molecule .
the tunnel splitting between these states is of order 100 hz ( in frequency units ) for 8 and unobservably small for 12 .
it must be stressed that in the solid , the tunneling is not of the coherent flip - flop type seen in the nh@xmath2 molecule , and previous authors have examined various decoherence processes by which the tunneling dynamics of a single molecule change from coherent to incoherent @xcite .
this is not enough to explain the observed non - exponential time behavior , for if there were a single characteristic time scale for relaxation of a single molecule , and all molecules relaxed independently , the magnetization would relax essentially exponentially in time with the same time scale as for one molecule .
thus , the non - exponential time behavior is a strong indicator that the molecules in the solid do not relax independently of each other .
the biggest and most obvious coupling between molecules is the dipole - dipole interaction , and while this has been considered by many previous authors @xcite a complete theory is still lacking .
in particular , the @xmath3 form has been previously explained by prokofeev and stamp @xcite , but as noted by gatteschi , sessoli , and villain @xcite it is unclear if it applies to all situations .
these latter authors also give a heuristic argument for the @xmath3 law for the particular case of the demagnetization problem .
we comment further or this below . in this paper , we report on our attempt to solve this problem .
our first approach is monte carlo simulation . in this
we follow in the footsteps of refs .
@xcite , and many aspects of our simulation and the results are very similar to those found by these authors .
we then try to understand our monte carlo results by developing a set of rate equations .
these rate equations entail the distribution of dipole fields at the molecular sites . for the specific problem of demagnetization
, we can construct an approximate model for this distribution , which then enables us to solve the rate equations numerically .
we find that the solution to the rate equations matches the monte carlo results quite closely .
furthermore , we can show analytically that that the solution obeys a square - root behavior with time at short times .
we emphasize that as in ref .
@xcite we have only studied the demagnetization problem .
further , the scaling behavior that we find for ancillary quantities also agrees entirely with ref .
thus , we can claim no priority for this result .
we also emphasize that our model for the dipole field distributions and the rate equations requires no further ingredients or fitting parameters beyond those involved in specifying the monte carlo process . in this paper
we only look at the problem of demagnetization of a spherical sample with a cubic lattice in order to minimize the complications from demagnetizing fields , and focus on the shape independent aspects of the problem , but we believe that our rate equation approach offers a method to attack a much wider class of problems , and in the future we hope to study other experimental protocols , sample shapes , and lattice types .
the plan of the paper is as follows . in
we describe the basic physical model underlying the relaxation @xcite .
we then describe our monte carlo simulations and results in .
the theory for the rate equations and the bias distribution are developed in secs .
[ theory ] and [ bias_dist ] .
finally , in we present our analytical solution to the rate equations , and the @xmath3 law .
as shown in ref .
@xcite , the fundamental process that governs the dynamical behaviour of the spins is as follows . in a short time interval @xmath4
, the spin of the @xmath5th molecule flips from @xmath6 to @xmath7 , or @xmath7 to @xmath6 , with a probability p_flip , i = _ i dt , with _
i ( e_i ) = 4 ^2 w -(e_i^22w^2 ) .
[ p_flip ] here , @xmath8 is defined via the statement that @xmath9 is the quantum mechanical amplitude per unit time for a spin to tunnel between the @xmath0 states , @xmath10 , where @xmath11 is the energy of dipole - dipole interaction between the molecular electronic spin and the nuclear spins of nearby nonmagnetic atoms such as n and h which are always present in the molecules studied , and @xmath12 is the energy of the @xmath13 state relative to the @xmath6 state due to the net magnetic field seen by the @xmath5th molecule
. we shall refer to @xmath12 as the bias on site @xmath5 @xcite . for 8 , @xmath14 k , @xmath15 mk , and @xmath16 k in temperature units .
( we shall set @xmath17 and @xmath18 to unity in all working formulas , so temperature , energy , and frequency all have the same units . )
the dominant feature in is the exponential suppression of the flip rate with the square of the bias energy @xmath12 , and a large part of this energy arises from the dipolar field of the other molecular spins in the solid , which can be estimated to be of order @xmath19 oe for near neighbour spins , leading to the energy scale @xmath20 k quoted above .
more explicitly , the dipolar part of @xmath12 is given by e_i , dip & = & _ ji k_ij _ j , [ ebias ] + k_ij & = & 2 e_dm a^3 r^3_ij ( 1 - 3z^2_ij r^2_ij ) .
[ kij ] here , @xmath21 is the energy scale of interaction for near neighbours , @xmath22 is the near - neighbour distance , @xmath23 is the distance between spins @xmath5 and @xmath24 , @xmath25 is the projection of the corresponding displacement onto the _
z_-axis , the easy axis of the spins .
finally , @xmath26 is an ising spin variable such that @xmath27 when the true spin on site @xmath5 is @xmath28 . since the dipole field is long ranged , and @xmath29 , the flip of the @xmath5th spin changes the bias field on a large number of neighbouring spins , and thus changes the flip probability for those spins significantly .
the relaxation of the magnetization of the entire solid is therefore a complex coupled process in which every individual spin essentially waits until it experiences a bias field less than @xmath30 in magnitude , and then flips with a probability per unit time equal to approximately @xmath31 .
the flip of this spin changes the bias field at many other molecules , and if one of them then happens to have a near - zero bias field , it flips , leading to the possibility of flips at yet more molecules .
@xcite refers to this scenario as a long - range glauber model .
as explained in , in this paper we only report on simulations on spherical samples of @xmath32 spins on a cubic lattice in order to eliminate the effects of inhomogeneous demagnetizing fields .
in addition , we only consider the demagnetization process .
thus , the spin @xmath26 is initialized to the value + 1 at every site .
starting from this configuration , we simulate the time evolution of the sample ( as described below ) for between 60 and 500 runs , and then average the total magnetization of the entire sample over these runs .
we have performed simulations for two sample sizes , with @xmath33 and 82,519 .
the initial spin polarization creates an almost delta - function - like distribution of bias fields centered at zero field , exactly as expected theoretically .
we see small deviations from a perfectly uniform distribution due to the finite size of the sphere .
the evolution of the system from time step @xmath34 to the next time step @xmath35 is carried out using the following protocol . at time @xmath34 , the bias energy @xmath12 is computed at every site using .
all spins are then flipped or not flipped using the flipping protocol described below .
we are now at time @xmath35 .
the bias fields are recomputed at all sites , and the process is repeated .
the flipping protocol we employ entails a slightly modified flip probability p_flip , i = _ 2 ^ 2 4w ( w - |e_i|)dt , [ p_flip2 ] instead of the original form ( [ p_flip ] ) . here
, @xmath36 is the heavyside function equal to unity for positive argument and to zero for negative argument . in other words
, a spin flips only if the bias field on it is less than @xmath30 in magnitude .
this modification is not material to the physics , and it reduces the run time of the simulations .
we refer to the spins in the window @xmath37 as _
reversible_. we have also used in a few cases , and not found any significant differences in the results .
further , @xmath38 , and the prefactor in is chosen to ensure that the integral @xmath39 is unchanged . in this way , the total magnetization that flips in a large subvolume containing many spins is unaffected . for future use
we define _ 0 = _ 2 ^ 2 4w .
an important consideration arises with regard to the values of @xmath21 , @xmath8 , and @xmath30 to be used in the simulation .
we know that the ratio of these quantities for real 8 is @xmath40 , and @xmath41 . due to the long ranged nature of the dipole field , when a spin flips , it has the potential to bring @xmath42 spins into the _ reversibility _
region @xmath37 .
we refer to this as the _ influence sphere _ of the spin . to overcome finite size effects
, we must make sure that our simulation includes a large number of influence spheres .
secondly , the rate at which a spin flips , even if it is within the reversibility window , is governed by @xmath8 , and our simulation would be much too slow if we used the actual value of @xmath43 .
we have therefore chosen different values for these quantities while still ensuring the physically important restriction @xmath44 . specifically , we take @xmath45 , @xmath46 , and vary @xmath30 over a range of values between @xmath47 and @xmath21 .
the next consideration is over the choice of the time step @xmath4 .
we set @xmath48 , and hence independent of @xmath30 .
this is done in order to remain true to the idea that the flip probability for a reversible spin should depend on @xmath30 only through the rate @xmath49 , and not @xmath4 . with our choice of @xmath4
this probability is p_flip & = & _ 2 ^ 2 4w dt + & = & 0.01 _ 2 ^ 2 4w e_dm _ 2 ^ 2 + & = & 0.01 e_dm 4w . by choosing @xmath50 ,
we ensure that the flip probability in one time step is not too large , which in turn ensures that our discretization of time is not too coarse , and that the simulation is sufficiently close to a continuous process . at the same time
, @xmath51 is large enough that we do not expend unnecessary time steps in waiting for the spin configuration to change by a meaningful amount .
the time - scale @xmath52 demarcates short versus long times , and we shall study relaxation for @xmath53 in some cases , i.e. , @xmath54 time steps . for real 8
we have @xmath55 secs .
some other details of the simulation are as follows .
the spherical sample is built from a cube having an odd number of sites on a simple cubic lattice with lattice constant ` a ' , and selecting those sites within a distance @xmath56 of the origin in order to get a sphere of diameter @xmath57 .
the two system sizes @xmath33 and @xmath58 correspond to sphere diameters @xmath59 and @xmath60 respectively .
the sites are indexed from 1 to @xmath32 , and their cartesian coordinates are stored in one - dimensional arrays . to reduce computer time , at the start a one - dimensional look - up table is made of the kernel @xmath61 by converting the triple of distances ( @xmath62 ) into a single unique number using some artificial but easy - to - implement formula that is invertible , i.e. , capable of yielding the triple ( @xmath62 ) from the single number . the central quantity of interest that is measured in our simulations is the magnetization , m = ( n _ - n_)/n , [ def_mag ] where @xmath63 and @xmath64 are the number of up and down spins .
the magnetization is measured at every time step .
in addition , we also measure at every time step , the bias distribution @xmath65 , defined such that @xmath66 is the fraction of spins experiencing a bias field between @xmath67 and @xmath68 .
the bin width for numerical purposes is chosen as @xmath30 itself as this is a sufficiently small number compared to @xmath21 .
secondly , the distribution is measured for biases that satisfy @xmath69 . in practice , we find that the fraction of sites that lie outside this range is @xmath70 .
as mentioned above , we have performed the simulations for different relative values of @xmath30 and @xmath21 . for a test case
, we made the contraphysical choice @xmath71 . in this case
we expect each spin to remain reversible most of the time , and rarely move out of the reversibility window when neighbouring spins flip .
each spin should then relax essentially independently of the others , leading to exponential relaxation of the magnetization with a rate @xmath72 .
this is indeed what is observed , giving us confidence in our numerical code .
the physically interesting simulations are performed for @xmath73 . in
, we show the magnetization versus time for one such simulation over a time @xmath74 .
it is evident that the decay of @xmath75 is nonexponential , and that there is a steep initial drop in @xmath75 over a time of order @xmath76 .
this drop is shown in more detail in , and is quite well fit by a square - root form ; we discuss this in more detail in . in both these figures
, we have performed an average over 60 runs .
( color online ) short - time behavior of the magnetization , with the same parameters as .
also shown is the result from numerical solution of the rate equations . ]
( color online ) short - time behavior of the magnetization , with the same parameters as .
also shown is the result from numerical solution of the rate equations . ] in we show the short- and long - time bias distribution @xmath65 for the same parameters as in figs . [ m_vs_t_long.eps ] and [ m_vs_t_short.eps ] . at short times , the distribution is marked by three clear peaks , as well as a few shoulders , which we shall explain in more detail in .
here we note that the two main peaks other than at the center are at @xmath77 and @xmath78 .
it is also evident that the peaks and shoulders become less distinct as @xmath34 increases .
indeed , for @xmath79 , they disappear completely , as shown in . here
we see a new feature developing , namely a hole in the distribution at @xmath80 , for @xmath81 .
( color online ) same as but for long times , and averaged over 10 runs only . note the reduced scale on the axis . ]
( color online ) same as but for long times , and averaged over 10 runs only . note the reduced scale on the axis . ]
the bias distributions also provide a good indicator of whether our system size is large enough and whether the averaging procedure is valid . to this end
, we show in the short- and long - time distributions for the smaller sample size ( @xmath82 ) but all parameters the same as in previous figures .
the two figures are drawn for averages over 60 and 30 runs , respectively . as can be seen , the statistical scatter is only minimally greater , and the quantitative features heights and locations of the peaks at short times , the hole at zero bias at long times are identical .
finally , in , we show the short - time distribution for a single run of the larger sample .
the features seen in the 60-run average are all clearly present , showing that questions of self - averaging do not arise in this system .
( color online ) same as for the @xmath82 sample .
the average is over 30 runs . ]
( color online ) same as for the @xmath82 sample .
the average is over 30 runs . ]
( color online ) single - run short - time bias distribution for the @xmath83 sample , with the same parameters as before .
there is no averaging . ]
to understand our simulations , we have developed a theory based on rate equations .
the key realization lies in the very different role played by the reversible and the nonreversible spins , and that we therefore need to understand the time - development of each set separately .
we denote by @xmath84 , @xmath85 , and @xmath86 the total number of reversible spins at any instant ( i.e. , those with a bias satisfying @xmath87 ) , and the parts of this number whose spins are up or down .
corresponding lower case symbols @xmath88 , @xmath89 , @xmath90 etc
. are used for the fractions @xmath91 , @xmath92 , @xmath93 , etc .
we also denote the number of nonreversible spins , @xmath94 , by @xmath95 , and the sets of spins of various types by @xmath96 , @xmath97 , @xmath98 etc .
these sets obey obvious relations such as @xmath99 and so on , which need not be listed .
it also pays to introduce the _ reversible magnetization _ , m_r = n_r - n_r , the total magnetic moment @xmath100 , and its reversible part , @xmath101 .
we now examine how different spins can develop in a small time interval @xmath4 . a non - reversible spin ( at site _
i _ , say ) can 1 .
_ move into _ the reversible bias range with a probability @xmath102 .
2 . _ remain _ in the non - reversible range with a probability @xmath103 .
+ naturally , since this spin can not flip in the interval @xmath4 , these possibilities depend on the behavior of other spins .
we shall address the probability @xmath102 below . a reversible spin ( again taken to be at site _
i _ ) , on the other hand , can do the following : 1 .
_ flip _ and _ move out _ of the reversible range with probability @xmath104 .
flip _ and _ remain _ in the reversible range with probability @xmath105 .
3 . _ not flip _ and _ become nonreversible _ with probability @xmath106 .
4 . _ not flip _ and _ stay reversible _ with reversibility @xmath107 . once again
, the probability @xmath108 depends on the behavior of other spins , and will be estimated below .
we have also introduced the quantity = _ 0 dt , in which the index _
i _ is omitted in @xmath109 , since this is the flip probability for _ all _ reversible spins .
clearly , our model assumes that the processes of flipping and of moving in or out of the reversibility range are independent , which in turn means that different spins flip or do not flip completely independently of each other , with a probability that depends only on the local bias .
this assumption will be valid provided the bias distribution @xmath65 is reasonably spatially homegeneous across the sample at all times .
such is the case for our spherical samples , but will need to be reexamined for other shapes . with the above hypothesis , the change in the numbers of various types of spins in a short time interval @xmath4 are easily written down . for @xmath110 , we have dn_r = -_i + _ i ( 1 - ) + _ i . the first four terms on the right correspond to the four processes enumerated above for reversible spins , while the fifth term corresponds to nonreversible up spins becoming reversible .
simplifying , we get dn_r = -_i + _ i ( 1 - ) + _ i . [ dnrup ] similarly , dn_r = -_i + _ i ( 1 - ) + _ i . [ dnrdn ] adding the last two equations , we get a very simple equation for the change in the total number of reversible spins , dn_r = -_i + _ i , [ dnr ] which does not depend on @xmath109 at all , since we do not discriminate between up and down spins in the set @xmath96 , and the changes in its size are a function of the behavior of neighboring spins of the members of this set . by taking the difference of
, we get the change in the unnormalized reversible magnetization : dm_r = dn_r - dn_r
. we can simplify the expression that results upon substitution of the actual forms of @xmath110 and @xmath111 by anticipating that the probabilities @xmath108 and @xmath102 will also be proportional to @xmath4 .
thus terms such as @xmath112 are @xmath113 and may be omitted . in this way , we get dm_r = -2m_r - ( _ i - _ i ) + ( _ i - _ i ) .
lastly , we find @xmath114 , the change in the total unnormalized magnetization .
since this change can come about only by the flipping of reversible spins , and since each flip changes @xmath115 by 2 , dm & = & -2 _
i + 2 _ i + & = & -2 m_r .
[ dm ] for the equations for @xmath118 , @xmath119 , and @xmath114 to be useful , we need the probabilities @xmath116 and @xmath117 .
let us begin by considering a nonreversible spin @xmath120 that sees a bias @xmath121 .
for this spin to move into the reversible range , _ reversible _ spins at other sites will need to flip and alter the bias at site _
i _ to satisfy @xmath122 , where the prime indicates the bias after a time interval @xmath4 .
now , e_i = e_i + _ j k_ij d_j . [ bias_shift ] here , the site _
i _ is excluded from the sum , and @xmath123 is the change in the spin at site _
j _ in the time @xmath4 .
the requirement that @xmath124 implies that only a particular set of reversible spins determined by the geometry of the lattice and the form of the dipole kernel @xmath61 can be effective in making spin _
we shall refer to such spins as _ triggering _ spins . to estimate their number
we make the critical simplification that we may ignore simultaneous spin flips since such processes will have a very low probability proportional to @xmath125 , which may be neglected as @xmath4 is infinitesimal .
thus , in , we take @xmath126 for all but one distant reversible spin . taking this spin to be up , so that @xmath127 , we get e_i = e_i - 2k_ij ( ) , where the arrow in @xmath128 indicates that the distant spin is up .
the condition @xmath124 then implies that k_ij ( ) .
[ k_range_1 ] similarly , if the distant spin is down , we require k_ij ( ) .
[ k_range_2 ] we now find the number of sites for which the couplings @xmath61 lie in the range ( [ k_range_1 ] ) or ( [ k_range_2 ] ) . if we define k_1(e_i ) = ( e_i - w ) , k_2(e_i ) = ( e_i + w ) , [ k1k2 ] then these two ranges correspond to intervals @xmath129 $ ] , and @xmath130 $ ] in which @xmath61 must lie .
let us denote the numbers of sites in each interval by @xmath131}$ ] and @xmath132}$ ] .
we have n_[k_1 , k_2 ] = _ k_1^k_2
g(k ) dk , and similarly for @xmath132}$ ] , where @xmath133 is the density of dipole couplings found in ref .
that is , @xmath134 is the number of sites for which the coupling to a central site lies between @xmath135 and @xmath136 .
we have g(k ) = , = 16 9 .
it then follows that n_[k_1 , k_2 ] & = & n_[-k_2 , -k_1 ] + & = & e_dm(1k_1 - 1k_2 ) [ nk1k2 ] + & = & 4 .
[ nk1k2_final ] since the number of distant sites at which a triggering spin could be located is independent of whether that spin is up or down , we can calculate the probability that spin @xmath5 will become reversible , that is to say @xmath102 , as the product of three factors : ( i ) the number ( [ nk1k2_final ] ) , ( ii ) the fraction of these sites at which the spin is itself reversible , @xmath88 , and ( iii ) the probability that any one of these spins will flip , @xmath109 .
thus , = 4 n_r e_dm w e_i^2 - w^2 _ 0 dt .
[ pin_i ] the above calculation assumes once again that the local reversible fraction @xmath88 in the vicinity of spin _
i _ is spatially homogeneous , and thus independent of the location of site _
we next turn to the calculation of @xmath117 , which proceeds in close parallel to that of @xmath116 . consider a reversible spin at site _ i _
, i.e. , the bias @xmath12 obeys @xmath137 .
we refer to this site as the central reversible spin .
this spin will become nonreversible if a distant reversible spin flips in such a way as to push the bias at site _
i _ outside the interval @xmath138 $ ] .
suppose the distant spin flips from up to down .
since we have @xmath137 and want @xmath139 , the coupling @xmath61 must be such that k_ij , where @xmath140 and @xmath141 are as defined in .
noting that now @xmath142 and @xmath143 , the number of sites that meet this requirement is given by _ -^k_1
g(k)dk + _ k_2^
g(k ) dk = 4 .
[ n_dist_up ] similarly , if the distant spin flips from down to up , the condition on @xmath61 is k_ij , which is met by a number of sites equal to _
-^-k_2 g(k)dk + _
-k_1^ g(k ) dk = 4 , [ n_dist_dn ] which is the same as .
thus , once again , the number of sites on which a triggering spin can be located is independent of whether that spin is up or down , and we may calculate @xmath108 as the product of ( i ) the number of sites ( [ n_dist_up ] ) , ( ii ) the fraction @xmath88 that the spin on one of these sites is reversible , and ( iii ) the probabibility @xmath144 that this spin will indeed flip .
thus , = 4 n_r e_dm w w^2 - e_i^2 _ 0 dt .
[ pout_i ] the expressions ( [ pin_i ] ) and ( [ pout_i ] ) suffer from unpleasant singularities when @xmath145 .
these singularities are unphysical , and are a consequence of using the modified spin - flip probabillity ( [ p_flip2 ] ) with the hard cutoffs at @xmath146 .
better estimates are obtained by noting that for @xmath102 , @xmath147 is likely to be much bigger than @xmath30 , while for @xmath108 the converse is true .
we therefore neglect the term @xmath148 in the denominator of and @xmath149 in the denominator of , leading to the expressions ( e_i ) & = & 4 n_r e_dm w e_i^2 _ 0 dt , [ pin ] + ( e_i ) & = & 4 n_r e_dm w _ 0 dt . [ pout ]
we note here the intuitively reasonable fact that @xmath117 is much greater than @xmath116 .
the set @xmath96 is much smaller than @xmath98 , so an initially reversible spin will be knocked out of reversibility by almost all flips of neighboring spins . by contrast , to move an initially nonreversible spin into reversibility , one must cancel the preexisting bias at the nonreversible site nearly exactly , which can only be done by flipping distant spins at a very specific set of sites . for
this same reason , @xmath108 essentially does not depend on @xmath12 , while @xmath102 does .
note also that @xmath116 and @xmath117 are both proportional to @xmath4 as anticipated earlier .
we now substitute into eqs .
( [ dnr ] ) , ( [ dmr ] ) , and ( [ dm ] ) for @xmath118 , @xmath119 , and @xmath114 , and divide by the total number of spins @xmath32 at the same time in order to get equations for intensive quantities .
let us begin by considering the two sums in one by one .
since @xmath108 is independent of @xmath12 as noted above , we have _
i = = 4 n^2_r e_dm w _ 0 dt . for the second sum
, we need to sum over the set @xmath98 .
we do this by including all sites where the bias exceeds @xmath30 in magnitude .
this leads to the approximation _
i = 4 n_r e_dm w f _ 0 dt , where @xmath150 is a dimensionless functional of the bias distribution @xmath65 , given by [ ( e ) ] = w^2 _ |e| > w ( e ) e^2 de . [ deff ] hence , = - 4 _ 0 e_dm w n_r ( n_r - f ) .
[ dnrdt ] next , we examine for @xmath151 . for the term with the sums over the sets @xmath97 and @xmath152 , we have , ( _ i - _ i ) = 4 n_r m_re_dm w _ 0 dt . for the remaining two sums
, we estimate the sizes of the sets @xmath153 and @xmath154 as @xmath63 and @xmath64 times the size of @xmath98 on the theory that when @xmath155 , most of the spins are nonreversible and the bias at any site is uncorrelated with whether the spin at that site is up or down , and that when @xmath156 , @xmath157 .
it follows that ( _ i - _ i ) 4
n_r m e_dm w f _ 0 dt . hence , = - 2_0 m_r - 4 _ 0 e_dm w n_r ( m_r - m f ) .
[ dmrdt ]
lastly , we obtain the equation for @xmath158 , which is the simplest of all : = -2_0 m_r .
[ dmdt ] equations ( [ dnrdt ] ) , ( [ dmrdt ] ) , and ( [ dmdt ] ) are the desired rate equations .
they are manifestly nonlinear , but more importantly and contrary to our initial hope , they are not a closed system because of the presence of the functional @xmath150 of the full bias distribution @xmath65 . at present
this puts a big limitation on their use . for the relaxation problem
we have been able to circumvent this limitation by constructing an interpolation form for @xmath65 which we believe is reasonably accurate and self - consistent over a wide range of times , well past that over which the square - root time development is seen .
we describe our approximation for @xmath65 in the next section .
as seen from the monte carlo simulations , the bias distribution at short times is dominated by three peaks at @xmath159 , @xmath160 , and @xmath161 .
the locations of the two side peaks are a strong indicator of their origin .
consider a site with its six nearest neighbors .
four of these neighbours are in the plane , and two are along the axis .
if any of the neighboring spins in the plane flips from up to down , the bias at the central site will change by an amount @xmath77 , while if any of the axis neighbours flips , the field at the central site will change by @xmath78 .
this explains the peak locations .
further , since there are twice as many near neighbours of any site in the plane as there are along the axis , we should expext the peak at @xmath77 to be about twice as high as the peak at @xmath78 as long as @xmath162 .
this is also seen in the data .
the smaller peak at @xmath163 and shoulder at @xmath164 can also be associated with spin flips at pairs of near neighbour sites .
motivated by this idea , we try and represent @xmath65 as a sum of three gaussians centered at 0 , @xmath165 , and @xmath78 .
suppose that at a given time , @xmath64 spins have flipped where @xmath166 , allowing us to ignore the possibility that two flipped spins are near neighbours of each other or even of a common third spin .
then there are @xmath167 spins that have a flipped neignbour in the plane , and @xmath168 spins that have a flipped neigbour along the axis , leaving @xmath169 spins which have no flipped neighbours at all .
thus the weights of the 0 , @xmath77 and @xmath170 peaks are proportional to @xmath171 , @xmath172 , and @xmath173 respectively .
we can further argue that the widths of all three peaks are equal and proportional to @xmath174 , since the fields at sites far away from all flipped spins should continue to vanish on average , but should have a variance that grows linearly with the number of flipped spins .
for a site next to a flipped spin , this variance is simply realized around the shift produced by the flipped neighbour .
thus for @xmath175 , the three - gaussian approximation ( tga ) to @xmath65 takes the form ( e ) ( 1 - 6 n _ ) g_0(e ) + 2 n _
g_+(e ) + 4 n _
g_-(e ) , [ rho_small_n ] where ( with @xmath176 , @xmath177 , or @xmath178 , and @xmath179 , @xmath180 , and @xmath181 ) g_(e ) = ( 2n _ ^2)^-1/2 e^-(e - e_)^2/2 n _ ^2 , ( n _ 1 ) .
[ g_small_n ] the quantity @xmath182 is @xmath21 times an unknown constant of order unity .
the arguments underlying start to become questionable for @xmath183 as small as @xmath20 , since sites with two near neighbour flipped spins start to become significant . to enable us to consider larger values of @xmath183
, we generalize the tga to the form ( e ) a_0 g_0(e ) + a_+ g_+(e ) + a_- g_-(e ) , [ rho_gen ] where g_(e ) = ( 2 ^ 2)^-1/2 e^-(e - e_)^2/2 ^2 .
[ g_gen ] that is , the peaks of the three gaussians are still taken to be at @xmath184 , @xmath185 and @xmath186 , the widths are taken to have a common value @xmath187 not necessarily proportional to @xmath174 , and the weights @xmath188 , @xmath189 , and @xmath190 are allowed to become arbitrary .
we will determine these weights and the width by the procedure described in the next subsection .
the form ( [ rho_small_n ] ) at small @xmath183 will serve as a check on the procedure .
it is apparent that the tga is qualitatively incapable of accounting for the very narrow hole that is burned in the distribution at long times , but here a different approximation scheme can be developed as the origin of the hole is physically obvious .
our discussion above implies that for very small @xmath183 , the flipped spins are randomly distributed in the lattice without any spatial correlations .
we therefore extend this idea to larger @xmath183 and consider a model in which the spin on each site is up or down independently of other spins , with probabilities @xmath191 , where @xmath75 is the magnetization .
we then calculate the first three moments of this model , and match those to the moments of the tga , .
these three moments , plus the normalization ( or zeroth moment ) give us the four conditions needed to determine the four quantities @xmath188 ,
@xmath190 , @xmath189 , and @xmath187
. the bias at any site is given by e_i = _ ji k_ij _ j. consider first the uniform spin configuration with @xmath192 , i.e. , @xmath193 for all _ i_. we know that in this case the bias vanishes at all sites except those in a narrow layer near the surface of our spherical sample .
hence we may take _
ji k_ij = 0 [ sumk ] for essentially all sites .
this result will be employed repeatedly in the calculations of the moments for configurations in which @xmath194 .
thus , for the first moment , we have & = & _ ji k_ij + & = & _ ji k_ij m + & = & 0 . [ avge ] similarly , for the second moment , we get = _ j , k k_ij k_ik .
the prime on the sum signifies that @xmath195 and @xmath196 .
now @xmath197 equals 1 if @xmath198 , and @xmath199 if @xmath200 .
hence , & = & _ j , k k_ij k_ik [ _ jk + ( 1 - _ jk ) m^2 ] + & = & _ ji k^2_ij ( 1 - m^2 ) + m^2 _ j k_ij _ k k_ik + & = & _ 2 e_dm^2 ( 1 - m^2 ) , [ avge2 ] where we have used , and defined _ 2 = 1e_dm^2 _ j i k^2_ij . numerical evaluation of the sum gives _ 2 = 53.427 [ kap2 ] for the third moment , we have = _ j , k , l k_ij k_ik k_il .
again , the prime signifies that @xmath195 , @xmath196 , and @xmath201 .
the only issue requiring care in performing the sum is the enumeration of the various cases of equality or inequality of the indices @xmath24 , @xmath202 , and @xmath203 . the first case is where all three indices are distinct .
then @xmath204 , and the contribution of this case to @xmath205 can be evaluated as _ 1 & = & m^3 _ j , k , l k_ijk_ikk_il ( 1 - _ jk ) ( 1 - _ kl ) ( 1 - _ lj ) + & = & m^3 _ j , k , l k_ijk_ikk_il ( 1 - 3_jk + 3 _ jk _ jl - _
jk_kl_lj ) + & = & m^3 + & = & 2m^3 _ ji k^3_ij .
[ ecube1 ] in line 2 above we have used the symmetry of the summand , and in line 4 we have used . the second case is where two of the indices @xmath24 , @xmath202 , and @xmath203 are the same , but distinct from the third .
now @xmath206 .
this case has three identically contributing subcases , and for its net contribution to @xmath205 we have _ 2 & = & 3 m _
j , k , l k_ijk_ikk_il _ jk ( 1 - _ jl ) + & = & 3 m + & = & -3 m _ ji k^3_ij , [ ecube2 ] where we have again used in the last line . the third and last case is that where @xmath207 .
now @xmath206 , and the contribution to @xmath205 is , therefore , _
3 = m _ ji k^3_ij .
[ ecube3 ] adding together eqs .
( [ ecube1 ] ) , ( [ ecube2 ] ) , and ( [ ecube3 ] ) , we get = -2 m ( 1 - m^2 ) _ ji k^3_ij .
we write this as = _
3 e_dm^3 m ( 1 - m^2 ) , [ avge3 ] where _ 3 = -2e_dm^3 _ ji k^3_ij = 190.47 , [ kap3 ] and the last result is found numerically .
it should be noted that in this model , the moments of @xmath67 are simply geometrical constants determined by the type of lattice times the appropriate power of the energy scale @xmath21 .
we now match the moments from the previous subsection with those of the three - gaussian approximation ( [ rho_gen ] ) .
the latter yields & = & 8e_dm a_+ - 4e_dm a_- , [ tgae1 ] + & = & ^2(a_0 + a_+ + a_- ) + 64 e_dm^2 a_+ + 16 e_dm^2 a_- , [ tgae2 ] + & = & 12 e_dm ^2 ( 2 a_+ - a_- ) + 512 e_dm^3 a_+ - 64 e_dm^3 a_- .
[ tgae3 ] equating these moments to those from the uncorrelated spin distribution yields 4 e_dm ( 2 a_+ - a_- ) & = & 0 , + ^2 ( a_0 + a_+ + a_- ) + 16 e_dm^2 ( 4 a_+ + a_- ) & = & _ 2 e_dm^2 ( 1-m^2 ) , + 12 e_dm ^2 ( 2 a_+ - a_- ) + 64 e_dm^3 ( 8 a_+ - a_- ) & = & _ 3 e_dm^3 m(1-m^2 ) . solving these equations along with the normalization condition , a_0 + a_+ + a_- = 1 , [ tgae0 ] we obtain a_0 & = & 1 - _ 3 128 m ( 1-m^2 ) , [ a0 ] + a_+ & = & _ 3 384 m ( 1-m^2 ) , + a_- & = & _ 3 192 m ( 1-m^2 ) , + ^2 & = & 14 ( 4_2 - _ 3 m ) ( 1-m^2 ) e_dm^2 .
[ sig_sq ] at this point let us ask whether the solution ( [ a0])([sig_sq ] ) approaches when @xmath175 . in that limit , since @xmath208 , @xmath209 .
feeding in the value @xmath210 , we get @xmath211 , and @xmath212 , instead of @xmath213 and @xmath214 .
the differences are rather small , however , and can be eliminated entirely if we make the replacement _ 3 _3 = 192 .
this leads to the final forms we shall use in our three - gaussian approximation , : a_0 & = & 1 - m ( 1-m^2 ) , [ a0fin ] + a_+ & = & m ( 1-m^2 ) , + a_- & = & m ( 1-m^2 ) , + ^2 & = & ( _ 2 - 48 m ) ( 1-m^2 ) e_dm^2 .
[ sig_sq_fin ] when we now compare the tga with the simulations , we discover that the agreement is off by @xmath215 if we use the value @xmath216 .
this value was calculated for an infinite lattice , and for a finite sized sample the variance of @xmath149 should be smaller . using the value 50 appropriate to the 82519 spin sample , we find that the agreement is considerably improved . in
we show the tga with the choice @xmath217 along with the results of the simulations for the @xmath218 spin sample for @xmath219 , @xmath220 , and @xmath221 , where @xmath222 . at these three times , @xmath223 , @xmath224 , and @xmath225 .
the agreement becomes poorer for larger @xmath34 , and it is about as good as could be expected given how simple - minded the approximation is .
( color online ) comparison between the three - gaussian approximation ( tga ) to the bias distribution and the simulation results for short times .
the sample has @xmath83 spins , and all other parameters are as in previous figures . ]
in we show @xmath226 for short times from our simulations , and from solving the rate equations with the value @xmath216 .
as can be seen the general trend is the same , although the detailed agreement is only good to about 3% .
once again , the agreement is improved if we set @xmath217 , as shown in .
the same data are shown on a log - log plot in . as can be seen ,
both the simulations and the rate equation show a power law behavior , with the same exponent .
the best fit gives an exponent of 0.46 , which is very close to 0.5 as it would be for @xmath3 behavior .
we now show that this behavior can be understood analytically on the basis of our rate equations , and that this exponent does not depend on the choice of @xmath227 .
the first key point is that starting from a delta - function at @xmath228 , the bias distribution becomes broader than the reversibilty region at some ultra - short time when the fraction of flipped spins is still very small . from
, we find that for @xmath175 , ^2 a^2 e_dm^2 n _ , where @xmath229 .
thus @xmath230 only as long as @xmath231 , which is of order @xmath232 .
for such ultrasmall values of @xmath183 , @xmath233 , and @xmath234 , so the rate equation for @xmath88 simplifies to = 23 _ 0 e_dmw .
this has the solution @xmath235 , and so the condition that @xmath230 holds only for @xmath236 , where ~1a^2 ( we_dm ) ^4 is an ultra - short time scale of order @xmath237 .
it follows that there is a large range of times , @xmath238 , for which @xmath239 even though @xmath175 , i.e. , very few spins are flipped .
thus almost all the weight in the bias distribution is still in the central gaussian , i.e. , @xmath240 , and the dimensionless functional that determines the repopulation of the reversibility region can be approximated as = 2 w^2 _ w^ e^-e^2/2 ^ 2 e^2 de .
now , by integrating by parts , we get _
w^ e^-e^2/2 ^ 2 e^2 de & = & 1w e^-w^2/2 ^ 2 - 1 ^ 2 _
w^ e^-e^2/2 ^ 2 de + & = & 1w e^-w^2/2 ^ 2 - 1 ^ 2 . the last expression can be expanded in powers of @xmath30 , and we get ( 1 - + ) .
[ fvssig ] the second key point is that even though @xmath175 , almost all the spins have been knocked out of the reversibility region , i.e. , @xmath155 . to see this we again approximate the bias distribution by neglecting the weight outside the central gaussian , and setting @xmath241 , so n_r _
-w^w e^-e^2/2 ^ 2 de . expanding the integrand in powers of @xmath67 and integrating , we get n_r ( 1 - w^26 ^ 2 + ) .
[ nrvssig ] thus , to first order in @xmath242 , @xmath243 , and the difference is of higher order : n_r - f = w^2 ^2 .
we can express this in terms of @xmath88 itself by using .
we have n_r , so n_r - f = 2 n_r^2 .
the rate equation for @xmath88 then reads = -2 n_r^3 , [ nreqn ] where we have defined = 4_0 e_dmw = .
the integration of is elementary . since this equation only holds for @xmath244 , we can write the integral in the form = ( t + t^ * ) , where @xmath245 is a time of order @xmath246 .
we thus have an explicit solution for the time dependence of the reversible fraction : n_r(t ) = 11 ( t+t^*)^1/2 .
the other rate equations can now be solved as follows .
we have by definition , n_r = n_r + n_r , m_r = n_r - n_r .
since @xmath247 , the answer for @xmath248 is immediate : m_r n_r = 11 ( t+t^*)^1/2 .
the rate equation for @xmath75 now reads = -2_0 1(t+t^*)^1/2 .
the integration is again elementary . assuming that @xmath249
, we can write the result as m(t ) 1 - , [ mvst ] where _ 1/2 = 16 _ 0 ^ 2 = 1 _ 2 ^ 2 e_dm .
is the experimentally observed @xmath3 law .
as noted in , a very pretty heuristic argument for this result is given in ref .
these authors reach the same conclusion by arguing that @xmath250 must be of the order of the typical dipole field when the spins start flipping , and thus proportional to @xmath251 , where @xmath252 is the typical distance between reversed spins .
they then note that @xmath253 , so that @xmath254 .
they then estimate @xmath255 as @xmath256 , from which it follows that @xmath257 , and that @xmath258 . as part of this argument
, one has that @xmath259 and that @xmath260 .
we find the same behavior for these quantities , but we arrive at it in a different ( and more difficult ! ) way since we did not have enough confidence in our understanding of the relation between @xmath250 and @xmath261 . instead , we find the delicate noncancellation between @xmath88 and @xmath150 in order to first find the differential equation obeyed by @xmath262 , and determine that @xmath260 , after which the equation for @xmath226 is elementary .
the agreement with @xcite gives us encouragement that our procedure is correct , and that we can use our more detailed rate equations to analyze other experimental protocols in the future .
99 c. sangregorio , t. ohm , c. paulsen , r. sessoli , and d. gatteschi , phys .
lett . * 78 * , 4645 ( 1997 ) .
t. ohm , c. sangregorio , c. paulsen , euro .
j. b * 6 * , 195 ( 1998 ) .
w. wernsdorfer , t. ohm , c. sangregorio , r. sessoli , d. mailly , and c. paulsen , phys .
lett . * 82 * , 3903 ( 1999 ) .
d. gatteschi , r. sessoli , and j. villain , _ molecular nanomagnets _
( oxford university press , oxford , 2006 ) .
this book gives a comprehensive and authoritative review of the entire field of smm s .
the problem of relaxation is especially ( but not exclusively ) discussed in chapter 9 .
l. thomas , a. caneschi , and b. barbara , phys .
lett . * 83 * , 2398 ( 1999 ) .
w. wernsdorfer , a. caneschi , r. sessoli , d. gatteschi , a. cornia , v. villar , and c. paulsen , phys .
lett . * 84 * , 2965 ( 2000 ) .
w. wernsdorfer , r. sessoli , a. caneschi , d. gatteschi , and a. cornia , europhys . lett . * 50 * , 552 ( 2000 ) . i. s. tupitsyn and b. barbara , in _ magnetism : molecules to materials _ iii , edited by j. s. miller and m. drillon ( wiley - vch , weinheim , 2002 ) . n. v. prokofev and p. stamp , phys .
lett . * 80 * , 5794 ( 1998 ) ; j. low . temp .
phys . * 113 * , 1147 ( 1998 ) .
a. cuccoli , a. fort , a. rettori , e. adam , and j. villain , euro .
j. b * 12 * , 39 ( 1999 ) .
j. f. fernandez and j. j. alonso , phys .
lett . * 91 * , 047202 ( 2003 ) ; _ ibid _ * 92 * , 119702 ( 2004 ) .
n. v. prokofev and p. stamp , j. low . temp .
* 104 * , 143 ( 1996 ) .
a. vijayaraghavan and a. garg , phys , rev .
b 79 , 104423 ( 2009 ) .
it is apparent that the bias @xmath263 , where @xmath264 is the magnetic moment of a molecule , and @xmath265 is the magnetic field at the @xmath5th site .
we prefer to work with the bias as all interactions in the problem are then expressed in terms of energies .
log - log plot of the short - time behavior of the magnetization . also shown
is the solution given by the rate equations with @xmath266 , and a linear fit to the latter .
this fit gives an exponent equal to 0.46 , close to 0.5 for an exact square root . ] | the low temperature relaxation of the magnetization in molecular magnetic solids such as 8 is studied using monte carlo simulations .
a set of rate equations is then developed to understand the simulations , and the results are compared .
the simulations show that the magnetization of an intially saturated sample deviates as a square - root in time at short times , as observed experimentally , and this law is derived from the rate equations analytically . |
supersymmetric yang - mills theories in @xmath1 dimensions are described by an action of the form @xmath2 where indices @xmath3 correspond to the adjoint representation of some semisimple compact lie algebra , @xmath4 are spacetime vector indices ( spatial indices will be denoted by @xmath5 ) , and @xmath6 are spinorial spacetime indices ; furthermore , the nonabelian field strength is formed from the vector potential @xmath7 via @xmath8 @xmath9 being the gauge group structure constants , and the gauge covariant derivative is given by @xmath10 the equality of the number of physical bosonic and fermionic degrees of freedom required by supersymmetry furthermore forces the number of transversal modes of the gauge field @xmath11 , @xmath12 , to be a power of two , as is the size of the corresponding irreducible spinor representation of the lorentz algebra .
one finds that it is in fact possible here to implement supersymmetry in @xmath13 . by truncation to configurations that have no spatial dependency ( i.e. dimensional reduction to zero space dimensions , see e.g.@xcite ) , one obtains supersymmetric quantum mechanics @xcite with a hamiltonian @xmath14 with @xmath15 .
the @xmath0 case with gauge group @xmath16 regained a lot of popularity after initial work by de wit , hoppe , and nicolai @xcite who showed the relevance of this particular model for the description of the eleven - dimensional supermembrane ( where @xmath16 appears as a regularized version of the lie group of area - preserving membrane diffeomorphisms ) through the so - called m(atrix ) theory conjecture @xcite , which states that this hamiltonian of a system of @xmath17 @xmath18-branes of type iia string theory should give a complete , non - perturbative description of the dynamics of @xmath19-theory in the light cone frame .. furthermore , these matrix models play an important role in the ikkt model @xcite , which may provide a non - perturbative description of iib superstring theory .
gauge groups of other types are also of interest here , as realizations of this model with @xmath20 and @xmath21 symmetry are given by systems of @xmath17 type - iia @xmath18-branes moving in orientifold backgrounds , cf .
@xcite .
one question of chief importance is that of normalizable zero energy vacuum states of these models ; this is difficult to settle as the potentials have flat valleys that extend to infinity , and hence , the corresponding ground states are at threshold . while it is exceedingly difficult to try to explicitly solve the schrdinger equation for these models , it is already of great interest to know the _ number _ of such normalizable ground states ; for example , it is of crucial importance to the matrix theory conjecture that there is exactly one such state for every @xmath17 in the models derived from @xmath16 gauged @xmath22 sym ( which just corresponds to a bound state of @xmath17 @xmath18-branes that appears as a graviton with @xmath17 units of momentum in the compactified direction ) .
while this is widely believed to be the case by now , the situation is still much less clear for other gauge groups .
as is nicely explained in @xcite , the number of normalizable ground states is given as the low temperature limit of the partition function @xmath23 but as the calculation of this quantity seems beyond reach for most systems of interest , it appears more promising to try to calculate the witten index @xmath24 instead ( where @xmath25 is the fermion number ) , as this should also give the number of normalizable ground states . in case of a discrete spectrum , this partition function would be @xmath26-independent , so we could as well take the limit @xmath27 , which is accessible in a perturbative calculation . for a continuous spectrum of the hamilton operator
, this does not work in general , since supersymmetry still pairs bosonic and fermionic modes , but the spectral density of scattering states need not be equal . in this case , the ` boundary ' term @xmath28 in the decomposition @xmath29 can acquire a nonzero value . while the technique of splitting the integral into a ` principal contribution ' from the bulk term as well as a ` deficit contribution ' which takes the form of a boundary term @xcite works remarkably well in many situations , as one frequently finds that the boundary term is zero even if it _ a priori _ does not seem to have to be ( e.g. @xcite ) , this is not the case in the systems at hand .
nevertheless , the boundary term has been calculated for @xmath30 in @xcite , and reasons ( that are based on the ( heuristic ) assumption that for the calculation of this deficit term the @xmath18-branes can effectively be treated as identical freely propagating particles , as in @xcite ) have been given in @xcite that this deficit term should be @xmath31 for the @xmath22 models derived from @xmath0 @xmath16 gauged sym , while one expects the value @xmath32 for the @xmath33 and @xmath34 models derived from @xmath35 , resp .
@xmath36 @xmath16 gauged sym . employing the mass deformation method that has been developed in @xcite , kac and smilga @xcite showed via group - theoretical means that under the hypothesis that no large mass bound state becomes non - normalizable as the zero - mass limit is taken , the number of normalizable ground states should be given by certain simple partition functions : @xmath37 independent arguments that lead to the same @xmath20 and @xmath21 multiplicities have been presented in @xcite which are based on an analysis of the hilbert space of a chiral fermion that is constructed from d0 branes at an orientifold singularity .
the bulk term @xmath38 is at least in principle independently accessible via a generalization @xcite of the brst deformation method which
was devised by moore , nekrasov , and shatashvili @xcite in order to greatly simplify the calculation of the corresponding partition functions by adding terms to the action which break all but one supersymmetry ( so that the partition functions do not change ) sym with low - rank gauge groups that this method seemingly also works for other groups besides @xmath16 ] .
concerning the boundary term , the method of @xcite has been generalized in @xcite to other gauge groups , but it was found that the expected witten index @xmath40 could not be obtained that way @xcite , indicating a failure of the assumption of the validity of the free hamiltonian approach used in @xcite .
nevertheless , for all the @xmath41 cases investigated previously , the expected vacuum degeneracies support the hypothesis @xcite that the boundary term is ( with the possible exception of @xmath42 ) always a small number in the interval @xmath43 $ ] , and hence a prediction of the witten index should be possible from the bulk index alone . a first evaluation of @xmath44 for special orthogonal and symplectic groups employing the moore method has been performed in @xcite
; there as well as in further work on @xmath45 @xcite , the algebraic bulkiness of bulk witten index calculations was pointed out , and indeed , even for the @xmath45 case , going to ranks far beyond @xmath46 already required specialized term manipulation code to be written . in the following ,
we want to review the operational issues of the application of the moore method to the even far more involved @xmath0 case and present techniques ( some conservative , some speculative ) that allow the calculation to be taken to regular simple groups of rank 4 and 5 , and also give arguments that such a direct approach is barely feasible for simple groups of rank @xmath47 employing currently available computer technology .
the heat kernel calculation @xcite of the bulk contribution to @xmath48 requires evaluation of the partition function @xmath49[a_\mu , a_\nu]+\frac{1}{2g^2}\tr\psi_\alpha[\gamma^\mu_{\alpha\beta}x_\mu,\psi_\beta]\right)\\ \displaystyle = \int\prod_{a=1}^{{\rm dim}\,g}\prod_{\mu=1}^d\frac{dx_\mu^a}{(2\pi)^{1/2 } } { \rm pf}\,(-if^{bcd}\gamma\cdot x^d ) \exp\left(\frac{1}{4g^2}\tr[a_\mu , a_\nu][a_\mu , a_\nu]\right)\\ \end{array}\ ] ] where the fermionic degrees of freedom have been integrated out , yielding a homogeneous pfaffian , cf .
these partition functions @xmath50 , which come from the reduction of sym to zero dimensions ( see e.g. @xcite ) have been dubbed ` yang - mills integrals ' .
the factor @xmath51 that relates it to the bulk witten index is basically the effective gauge group volume , see @xcite for details .
the brst deformation technique of @xcite greatly simplifies this to the calculation of the integral @xcite @xmath52 where @xmath53 is the lie group rank , @xmath54 is the quotient of the orders of the center and the weyl group ( @xmath55 for @xmath56 , @xmath57 for @xmath58 , @xmath59 for @xmath60 , 1/12 for @xmath61 ; these are the only cases we are concerned with here ) , @xmath62 are auxiliary real quantities with @xmath63 which the end result will not depend on ( at least for sufficiently generic values of @xmath64 where no ` accidental ' merging of poles / zeroes happens ) .
@xmath65 is the set of roots of the lie algebra @xmath66 .
actually , as it stands these integrals do not make much sense , as we have poles on the real axis of integration .
( furthermore , they do not fall off fast enough towards complex infinity to rigorously justify closing the integration contour . )
the correct ( not yet fully justified , see @xcite ) interpretation of these integrals rather is the following algorithmic one : 1 . successively eliminate all @xmath67 ` integration variables ' in the term @xmath68 where the @xmath69 are all treated as real , and @xmath70 picks up the residues with positive imaginary part only .
( all the @xmath67 except the one being integrated out are treated as real . ) 2 .
substitute @xmath71 3 .
evaluate at generic values of @xmath72 .
( alternatively : simplify to find that the result does not depend on @xmath72 . ) due to the large number of factors in the denominator that occur in these formulae for all but the smallest simple groups , one has to resort to employing clever tricks to simplify the calculation , or massive computer aid ( or both ) . while the use of symbolic manipulation programs like maple or mathematica suggests itself for calculations like the ones at hand , and
indeed interesting results have been obtained that way @xcite , the range to which these calculations can be carried by employing such systems in a head - on approach is quite limited , mostly due to the observation that ( unlike the @xmath35 case ) in the evaluation of @xmath73 integrals , one seems to generically encounter poles of order @xmath74 that require forming @xmath75th order derivatives of very large products , leading to an explosion of the number of terms generated .
nevertheless , it is obviously important to try to obtain these values for as many gauge groups as possible in order to check the validity of various assumptions that had to be made in the calculation of witten indices , and perhaps even make further conjectures about analytical expressions , e.g. as in @xcite for @xmath35 .
even if one can not do much about combinatoric explosion here , the question nevertheless arises , whether considering the conceptual simplicity of the problem
one may be able to go a few steps further by trying to make use of as much of the structure of the calculation as possible in a dedicated program .
while it is frequently possible to outperform general - purpose symbolic algebra packages by three orders of magnitude in term complexity with such an approach , more than one rabbit has to be pulled out of the hat in order to achieve a performance gain of one million or more , which we will see to be necessary here .
as the underlying techniques are of interest for a far larger class of symbolic calculations , yet not widely known since there is ( to the author s knowledge ) hardly any literature on the relevant issues , we want to briefly present some of the fundamental concepts one should be aware of when taking symbolic algebra to its limits .
incidentally , the calculation at hand is an almost perfect example to study these techniques . conventional symbolic algebra frequently ( and especially in calculations like these ) wastes most of its time generating intermediate quantities of very limited lifetime in dynamically allocated memory which has to be reclaimed soon after .
in the following , we want to adhere to the convention to call this allocation of dynamic memory which will not explicitly be reclaimed _
consing_. as unnecessary consing forces frequent expensive calculations of reachability graphs of in - memory objects ( in order to identify reclaimable space in the garbage collection process ) , it should as a rule of thumb be avoided where possible . furthermore , one should keep in mind that for the presently dominant computer architecture the processor s memory interface is an important bottleneck , and hence it makes sense to try to find tight memory encodings for those pieces of data on which most of the calculation operates so that cache stalls are minimized .
one particularly striking feature of the calculation at hand is that naively integrating out a single variable makes conventional term manipulation programs first allocate huge numbers of almost similar terms , which are then checked ( in an overwhelming number of cases in vain ) for possible annihilations .
this can be avoided by re - structuring the calculation in such a way that instead of generating and processing large amounts of individual mostly similar ( yet different enough that cancellations become quite rare ) terms , one and the same term backbone is destructively modified to consecutively represent every single new term generated and then do further processing on this backbone wherever possible .
this technique is particularly useful when it comes to processing higher - order derivatives ( where in addition we take care of performing the iterations over places of factors where to derive in such a way that re - occurring combinations of derivatives , as in @xmath76 are only generated once ) .
one particular refinement for the present calculation is that one generally can hardly avoid creating new terms by substitution , except in the very last step ( i.e. integrating out the last variable ) where this is indeed feasible .
since the last step is also executed most frequently , this almost buys us an extra rank for free .
( a slightly nontrivial subtlety for the @xmath0 calculation here is that we have to take care of the possibility of generating factors @xmath77 in both the numerator as well as the denominator .
it may well happen that while accumulating the factors of a term , one intermediately encounters more such ` powers of zero ' in the denominator than in the numerator . ) as calculations with exact numbers cause some systems to perform excessive number consing , it is wise to try to avoid the exact rational number data type for all those parts of a calculation where one can use with impunity more limited data types that have fast direct hardware support ( like 32-bit integers ) . for our present calculation
this means to represent the coefficients of linear functions as machine integers .
( clearly , one has to take care of proper handling of some ` balancing denominators ' that are generated by this somewhat artificial treatment that requires to take least common multiples at substitutions . ) for all of the problems that are reachable with present computer hardware , all of the substitutions of such terms among themselves will not lead us outside the range of these integer machine data types .
a further property of the calculations under study is that when performed in the way described here , a large fraction of the contributions are zero , and in many cases , this can be detected somewhat easily without doing costly multiplications .
hence , it makes sense to pre - scan a term for being zero before processing it wherever applicable .
it is perhaps one of the less obvious properties of these yang - mills integrals that one can do much better than to substitute particular values for @xmath78 subject to the constraint @xmath63 for the final evaluation ; by bringing all factors to a proper lexicographical normal form ( @xmath79 contribution first if present , then @xmath80 contribution , then @xmath81 contribution , all after having substituted out @xmath82 ) , one observes to obtain also the correct final value by replacing every linear term by its leading coefficient . ; this was discovered through an intermediate programming error . ] while the largest calculations at present can not be done without employing this trick , we do a cross - check using the more conventional method wherever this is possible .
this is indicated in the last column of our table of results . as a special refinement ,
one notes that in the final gathering of powers of linear factors , the most frequently encountered numerators and denominators are @xmath83 as well as small powers of 2 .
indeed , a further noticeable speedup can be achieved by treating these factors special ( remembering one overall power of 2 as well as the resulting sign ) in order to avoid unnecessary use of exact rational arithmetics . as a basis for the implementation , the objective caml system @xcite appears as very appealing , since it is highly portable to a variety of different platforms , contains an optimizing compiler that can generate compact standalone binaries , allows a very smooth and easy two - way integration of c libraries and code ; all these qualities are highly desirable here especially since the problem at hand suggests itself to massive parallelization , perhaps by making use of donated computation time .
( while rudimentary parallelization support is present in our implementation and has proven its usefulness , we do not implement such a large - scale scheme here , mainly because it is expected that one would need a prohibitively large number of volunteers to successfully do the next rank . ) a further advantage may be that ocaml code generally is perceived as much less alien by the uninitiated than lisp code .
one further noteworthy issue here is the quality of the built - in exact rational number arithmetics , since there are all but obvious huge performance differences between various implementations .
while the implementation present in ocaml 3.07 is slower than the one of clisp by a factor of roughly 300 , this is not a big problem , since our techniques to reduce the use of fraction arithmetics are powerful enough ( at least when making use of the evaluation shortcut described above ) to make the amount of time spent in arithmetics comparable to the time spent in other parts of the program .
clisp would perhaps hardly be a viable alternative despite its excellent arithmetics implementations , since it is only a byte - code interpreter system . as there are good reasons to make the code with which these calculations have been performed publicly available , it has been included in the arxiv.org preprint upload of this work .
the following table shows our new as well as the previously known values for the bulk contributions to the witten index .
numerical approximations as well as the expected values from the kac / smilga hypothesis are also given .
calculation times refer to accumulated wall clock times on a stepping-9 intel pentium 4 cpu , 2.4 ghz , hyperthreading enabled / ocaml 3.07 , gcc 2.95 , linux kernel 2.4.24 student computer pool installation .
( values obtained on other systems have been re - scaled appropriately . )
boldface indicates new values .
some features that deserve special attention here are the particularly large values for @xmath84 and @xmath85 that would require @xmath86 , the @xmath87 and @xmath88 values that would require @xmath89 , and the dramatic explosion of calculation time ; beyond some ` trivial ' cases , increasing the rank by one costs a factor ( as a rule of thumb ) of roughly 1000 in cpu time in the interesting regime .
hence , it is probably not yet feasible to try to attack rank @xmath90 for the @xmath73 groups .
as the @xmath91 bulk indices are just the expected ones , and as the @xmath73 are of the right magnitude , with denominators being powers of two ( this is not the case at all for individual summands ) , there is good reason to believe in both the validity of this approach and its implementation .
however , the calculation of the @xmath92 index by using explicit values for @xmath72 seems to fail systematically for yet unknown reasons , despite a very careful analysis of the code .
( this may be related to yet another bug in the ocaml compiler ; this assumption is also fueled by the observation that part of the @xmath93 calculation causes memory violations that should not be possible at all in pure ocaml for a very specific combination of processors and optimization flags . )
all in all , the new data clearly show that for other gauge groups than @xmath16 , the issue of the number of vacuum states is not well understood , and more work has to be put into the determination of witten indices . in particular , we do not have at present a useful theory to calculate the boundary contributions , and it is well conceivable that some of the assumptions behind @xmath94 may be violated .
i want to thank matthias staudacher for introducing me to this interesting problem , and also for many helpful comments and discussions , the administrative staff of cip.physik.uni-munchen.de , where some of the calculations have been performed , as well as steffen grunewald for letting me submit some calculation jobs to the geo600 merlin cluster at aei .
furthermore , i want to thank the debian ocaml package maintainer , sven luther , for his for his prompt handling of bug reports , and damien doligez for fixing an ocaml compiler bug @xcite that was discovered through this work
. l. alvarez - gaume , `` supersymmetry and the atiyah - singer index theorem , '' commun .
* 90 * ( 1983 ) 161 .
p. austing , `` yang - mills matrix theory , '' arxiv : hep - th/0108128 . m. baake , m. reinicke and v. rittenberg , `` fierz identities for real clifford algebras and the number of supercharges , '' j. math . phys .
* 26 * ( 1985 ) 1070 .
t. banks , w. fischler , s. h. shenker and l. susskind , `` m theory as a matrix model : a conjecture , '' phys .
d * 55 * ( 1997 ) 5112 [ arxiv : hep - th/9610043 ] .
m. claudson and m. b. halpern , `` supersymmetric ground state wave functions , '' nucl .
b * 250 * ( 1985 ) 689 .
u. h. danielsson , g. ferretti and b. sundborg , `` d - particle dynamics and bound states , '' int .
j. mod .
a * 11 * ( 1996 ) 5463 [ arxiv : hep - th/9603081 ] . b. de wit , j. hoppe and h. nicolai , `` on the quantum mechanics of supermembranes , '' nucl . phys .
b * 305 * ( 1988 ) 545 .
t. fischbacher , sven luther , + http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=224417 d. friedan and p. windey , `` supersymmetric derivation of the atiyah - singer index and the chiral anomaly , '' nucl .
b * 235 * ( 1984 ) 395 .
r. flume , `` on quantum mechanics with extended supersymmetry and nonabelian gauge constraints , '' annals phys .
* 164 * ( 1985 ) 189 .
m. b. green and m. gutperle , `` d - particle bound states and the d - instanton measure , '' jhep * 9801 * ( 1998 ) 005 [ arxiv : hep - th/9711107 ] .
a. hanany , b. kol and a. rajaraman , `` orientifold points in m theory , '' jhep * 9910 * ( 1999 ) 027 [ arxiv : hep - th/9909028 ] .
n. ishibashi , h. kawai , y. kitazawa and a. tsuchiya , `` a large - n reduced model as superstring , '' nucl .
b * 498 * ( 1997 ) 467 [ arxiv : hep - th/9612115 ] .
d. kabat and p. pouliot , `` a comment on zero - brane quantum mechanics , '' phys .
* 77 * ( 1996 ) 1004 [ arxiv : hep - th/9603127 ] . v. g. kac and a. v. smilga , `` normalized vacuum states in n = 4 supersymmetric yang - mills quantum mechanics with any gauge group , '' nucl .
b * 571 * ( 2000 ) 515 [ arxiv : hep - th/9908096 ] .
w. krauth and m. staudacher , `` yang - mills integrals for orthogonal , symplectic and exceptional groups , '' nucl .
b * 584 * ( 2000 ) 641 [ arxiv : hep - th/0004076 ] .
w. krauth , h. nicolai and m. staudacher , `` monte carlo approach to m - theory , '' phys .
b * 431 * ( 1998 ) 31 [ arxiv : hep - th/9803117 ] .
x. leroy , `` le systme caml special light : modules et compilation efficace en caml , '' inria research report nov 1995 , 2721 .
url : http://pauillac.inria.fr/~xleroy/publi/caml-special-light-rr.ps.gz ; the ocaml website is http://www.ocaml.org g. w. moore , n. nekrasov and s. shatashvili , `` d - particle bound states and generalized instantons , '' commun . math .
* 209 * ( 2000 ) 77 [ arxiv : hep - th/9803265 ] .
v. pestun , `` n = 4 sym matrix integrals for almost all simple gauge groups ( except e(7 ) and e(8 ) ) , '' jhep * 0209 * ( 2002 ) 012 [ arxiv : hep - th/0206069 ]
. m. porrati and a. rozenberg , `` bound states at threshold in supersymmetric quantum mechanics , '' nucl .
b * 515 * ( 1998 ) 184 [ arxiv : hep - th/9708119 ] .
s. sethi and m. stern , `` invariance theorems for supersymmetric yang - mills theories , '' adv .
theor . math .
phys . * 4 * ( 2000 ) 487 [ arxiv : hep - th/0001189 ] .
s. sethi and m. stern , phys .
b * 398 * ( 1997 ) 47 [ arxiv : hep - th/9607145 ] .
a. v. smilga , `` witten index calculation in supersymmetric gauge theory , '' nucl .
b * 266 * ( 1986 ) 45 [ yad . fiz .
* 42 * ( 1985 ) 728 ] .
m. staudacher , `` bulk witten indices and the number of normalizable ground states in supersymmetric quantum mechanics of orthogonal , symplectic and exceptional groups , '' phys .
b * 488 * ( 2000 ) 194 [ arxiv : hep - th/0006234 ] .
w. taylor , `` m(atrix ) theory : matrix quantum mechanics as a fundamental theory , '' rev .
* 73 * ( 2001 ) 419 [ arxiv : hep - th/0101126 ] . c. vafa and e. witten ,
`` a strong coupling test of s duality , '' nucl .
b * 431 * ( 1994 ) 3 [ arxiv : hep - th/9408074 ] .
yi , `` witten index and threshold bound states of d - branes , '' nucl .
b * 505 * ( 1997 ) 307 [ arxiv : hep - th/9704098 ] . | values for the bulk witten indices for @xmath0 yang - mills integrals for regular simple groups of rank 4 and 5 are calculated by employing the brst deformation technique by moore , nekrasov and shatashvili .
the results can not be reconciled with the double assumption that the number of normalizable ground states is given by certain simple partition functions given by kac and smilga as well as that the corresponding boundary term is always negative .
# 1tr#1 # 1tr#1 # 1res^+_#1 hep - th/0312262 + aei-2003 - 113 + * bulk witten indices from @xmath0 yang mills integrals * * t. fischbacher + * |
the question of galaxy rotation curves and cluster masses has been known to require some form of energy density that makes its presence felt only by its gravitational effects since @xcite analyzed the velocity dispersion for the coma cluster . by 1959
there were eight galaxy rotation curves available from radio observations demonstrating a `` flat rotation velocity '' . in @xcite
our study of galaxy rotation curves without non - baryonic dark matter we demonstrated good fits to our sample of 101 lsb , hsb and dwarf galaxies including 58 galaxies that were fit photometrically with the single parameter @xmath2 ( 29 @xmath3band and 29 @xmath4band ) .
as a follow - up , we apply here the same framework to the question of x - ray galaxy cluster masses @xcite . in this framework
, we apply a generalization of einstein s general relativity ( gr ) based on a pseudo - riemannian metric tensor and a skew symmetric rank three tensor field @xmath0 , called metric - skew - tensor - gravity ( mstg ) . a renormalization group ( rg ) framework @xcite for mstg was developed to describe the running of the effective gravitational coupling constant @xmath5 , and the effective coupling constant @xmath6 that measures the strength of the coupling of the @xmath0 field to matter .
a momentum cutoff identification @xmath7 associates the rg scales to points in spacetime . for the static , spherically symmetric solution ,
the rg flow equations allow a running with momentum @xmath8 and proper length @xmath9 for the effective newtonian coupling constant @xmath10 , the coupling constant @xmath11 , and the effective mass of the skew field @xmath12 where @xmath13 denotes the radial coordinate .
the form of @xmath14 as a function of @xmath13 , obtained from the modified newtonian acceleration law , leads to agreement with solar system observations , terrestrial gravitational experiments and the binary pulsar psr 1913 + 16 observations , while strong renormalization effects in the infrared regime at large distances lead to fits to galaxy rotation curves and x - ray cluster masses . in @xcite , a scalar - tensor - vector gravity theory is formulated , which yields the same weak field modified acceleration law as in mstg @xcite , and equations that describe an effective running of @xmath5 , @xmath6 and @xmath15 with space and time .
a fit to the bulk properties ( the gas mass ) of 106 x - ray galaxy clusters is achieved without exotic dark matter by a best fit result using a nonlinear least - squares fitting routine including estimated errors for the one free parameter in the rg flow equations , @xmath16 , which fully describes the large distance renormalization of newton s constant , @xmath17 .
the running of newton s constant , @xmath14 , is fully constrained by choosing the remaining free parameter in the rg flow equations , @xmath18 , to take on a value proportional to the scale of the galaxy cluster ( for small clusters with galactic scale icm gas masses ) or to take on a particular constant value for the majority of the regular size galaxy clusters ( where the icm gas mass dominates over the masses of the individual galaxies ) .
thus , the individual cluster mass profiles as functions of radial position throughout the range of x - ray observations are parameter free predictions ; and are consistent with x - ray observed gas mass profiles without exotic dark matter .
the galaxy cluster s bulk properties fit and the specific fit to the individual clusters are compared to those obtained using milgrom s phenomenological mond model which does not fit the x - ray galaxy cluster masses unless an auxiliary dark matter component is included @xcite . @xcite
were able to account for the mond discrepancy between the x - ray observationally determined gas mass and the dynamical mass of the coma cluster by increasing the mond acceleration by a factor of four greater than that used to fit the galaxy rotation curves .
@xcite present evidence from the central 200 kpc of three clusters implying that mond is inconsistent with the observed temperature gradient which inflates the discrepancy in the mond acceleration to a factor of @xmath19 10 .
more recently , @xcite use x - ray data from the xmm - newton satellite for eight clusters of varying temperature and masses to place constraints on the use of the mond phenomenology . without treating the mond acceleration as a free parameter as opposed to a universal constant , mond predicts dynamical masses greatly in excess of the x - ray observations necessitating the ad hoc addition of dark matter to explain away the missing mass .
we are able to show that there is no missing mass when applying the mstg acceleration law to galaxy clusters .
recent observations from the xmm - newton satellite suggest that the intracluster medium ( icm ) is very nearly isothermal inside the region defined by the x - ray emission with temperatures ranging from @xmath20 115 kev ( or @xmath21 @xmath22 k ) for different clusters @xcite . the combination of the observed density profile , @xmath23 , and the temperature profile , @xmath24 , obtained from x - ray observations of the galaxy cluster leads to a pressure profile , @xmath25 , which directly leads to a mass profile , @xmath26 , by assuming the gas is in nearly hydrostatic equilibrium with the gravitational potential of the galaxy cluster . within a few core radii , the distribution of gas within a galaxy cluster
may be fit by a king `` @xmath1-model '' .
the observed surface brightness of the x - ray cluster can be fit to a radial distribution profile @xcite : @xmath27^{-3 \beta + 1/2},\ ] ] resulting in best fit parameters , @xmath1 and @xmath28 .
a deprojection of the @xmath1-model of equation ( [ betamodel ] ) assuming a nearly isothermal gas sphere then results in a physical gas density distribution @xcite : @xmath29^{-3 \beta /2},\ ] ] where @xmath30 is the icm mass density profile . provided the number density , @xmath31 , traces the actual mass , we may assume that @xmath32 , which according to @xcite is explicitly @xmath33 and rewrite equation ( [ betarhomodel ] ) @xmath34^{-3 \beta /2}.\ ] ] for a spherical system in hydrostatic equilibrium , the structure equation can be derived from the collisionless boltzmann equation @xmath35 where @xmath36 is the gravitational potential for a point source , @xmath37 and @xmath38 are mass - weighted velocity dispersions in the radial ( @xmath13 ) and tangential ( @xmath39 ) directions , respectively . for an isotropic system , @xmath40 the pressure profile , @xmath25 , can be related to these quantities by @xmath41 combining equations ( [ cbe ] ) , ( [ isotropicsystem ] ) and ( [ pressureprofile ] ) , the result for the isotropic sphere is @xmath42 for a gas sphere with temperature profile , @xmath24 , the velocity dispersion becomes @xmath43 where @xmath8 is boltzmann s constant , @xmath44 is the mean atomic weight and @xmath45 is the proton mass .
we may now substitute equations ( [ pressureprofile ] ) and ( [ velocitydispersion ] ) into equation ( [ isotropicsphere ] ) to obtain @xmath46 performing the differentiation on the left hand side of equation ( [ isotropicsphere ] ) , we may solve for the gravitational acceleration : @xmath47.\end{aligned}\ ] ] for the isothermal isotropic gas sphere , the temperature derivative on the right - hand side of equation ( [ accelerationprofile ] ) vanishes and the remaining derivative can be evaluated using the @xmath1-model of equation ( [ betarhomodel ] ) : @xmath48
given an isotropic density distribution , @xmath30 , the associated mass profile is @xmath49 where @xmath26 is the total mass contained within a sphere of radius r. for the @xmath1-model of equation ( [ betarhomodel ] ) , we may approximate the integral of equation ( [ massprofile ] ) for @xmath50 and @xmath51 @xcite : @xmath52 this result clearly diverges in the limit as @xmath53 ; but galaxy clusters are observed to have finite spatial extent .
this allows an approximate determination of the total mass of the galaxy cluster by first solving equation ( [ betarhomodel ] ) for the position , @xmath54 , at which the density , @xmath55 , drops to @xmath56 , or 250 times the mean cosmological density of baryons : @xmath57 then , provided @xmath58 , @xmath59 the dynamical mass in newton s theory of gravitation can be obtained as a function of radial position by replacing the gravitational acceleration with newton s law : @xmath60 where @xmath61 is newton s `` bare '' gravitational constant so that equation ( [ accelerationprofile ] ) can be rewritten as @xmath62,\ ] ] and the isothermal @xmath1-model result of equation ( [ isothermalaccelerationprofile ] ) can be rewritten as @xmath63 similarly , the dynamical mass in mstg can be obtained as a function of radial position by substituting the mstg gravitational acceleration law @xcite : @xmath64 so that our result for the isothermal @xmath1-model becomes @xmath65 we can express this result as a scaled version of equation ( [ newtonsmass ] ) or the isothermal case of equation ( [ isothermalnewtonsmass ] ) : @xmath66\biggr\}^{-1 } m_{\rm n}(r),\end{aligned}\ ] ] where we have made explicit the form of the running of g(r ) as in @xcite : @xmath67\biggr\}.\ ] ] in the limit of large @xmath13 , we show in @xcite that @xmath68 and the total mass of the cluster in mstg can be computed by taking equations ( [ mstgmassscaled ] ) to the same limit : @xmath69 it is a simple matter to solve equations ( [ mstgmassscaled ] ) and ( [ mstgmass ] ) explicitly for @xmath70 and @xmath71 , respectively , by squaring both sides and subsequently applying the canonical solution to the quadratic equation .
the derivation in milgrom s phenomenological mond model @xcite follows the same procedure , but utilizes the mond gravitational acceleration law , described by @xmath72 where @xmath73 is a function that interpolates between the newtonian regime , @xmath74 , when @xmath75 and the mond regime , @xmath76 , when @xmath77 .
the function and critical acceleration normally used for galaxy and cluster fitting are , respectively , @xmath78 applying equation ( [ milgromacc ] ) to either equation ( [ accelerationprofile ] ) or the isothermal case of equation ( [ isothermalaccelerationprofile ] ) yields the mond dynamical mass in terms of the newtonian dynamical mass of equation ( [ newtonsmass ] ) or the isothermal case of equation ( [ isothermalnewtonsmass ] ) : @xmath79 it is a simple matter to solve equation ( [ mondmass ] ) explicitly for @xmath80 by squaring both sides and subsequently applying the canonical solution to the quadratic equation . unlike the newtonian and mstg isothermal spheres whose densities fall off as @xmath81 in the limit of large @xmath13 , @xcite showed that the mond isothermal spheres have densities that fall of as @xmath82 where @xmath83 at large radii . in the newtonian and mstg cases ,
the behavior is keplerian at large radii and so the isothermal spheres have masses which diverge with @xmath13 @xcite . thus according to equations ( [ newtonslaw ] ) and ( [ rung ] ) the acceleration at large radii for newtonian and mstg isothermal spheres goes as @xmath84 .
conversely , the mond isothermal spheres have convergent masses regardless of the spatial extent @xcite : @xmath85 the mond acceleration law regularizes the integral of equation ( [ massprofile ] ) whereas a cutoff is necessary in the newtonian and mstg case , corresponding to the finite spatial extent of galaxy clusters as shown in equation ( [ totalgasmass ] ) .
however , as clearly shown in @xcite , the x - ray surface brightness distribution of these mond isothermal spheres provides a poor representation of the observed surface brightness distribution .
this distinction will become apparent in the section ahead where we study the mstg and mond mass profiles of the coma cluster .
we have adopted the compilation of @xcite as our sample .
the relevant cluster properties are listed in table [ clusterproperties ] arranged as follows : column ( 1 ) lists the cluster name truncated to 8 characters .
column ( 2 ) is the observed x - ray temperature .
column ( 3 ) lists the icm central mass density , @xmath86 , of the icm in units of @xmath87 .
column ( 4 ) is the @xmath1 parameter .
column ( 5 ) is the core radius parameter , @xmath28 , in units of kpc assuming @xmath88 @xcite .
column ( 6 ) is the position , @xmath54 at which the density , @xmath55 , drops to @xmath56 , or 250 times the mean cosmological density of baryons .
column ( 7 ) through ( 9 ) lists the icm gas mass , @xmath89 , the newtonian dynamic mass , @xmath90 , and the mstg dynamic mass , @xmath71 , respectively , each integrated to @xmath54 .
column ( 10 ) lists the `` convergent '' mond dynamic mass . in order to calculate the mstg dynamic mass we first need to phenomenologically
determine the running of the parameters , @xmath16 , and @xmath18 this describes the running of newton s constant at the scale of clusters according to equation ( [ runningg ] ) .
however , unlike the case of the galaxy rotation curves where a satisfactory fit to lsb and hsb galaxy data is obtained with the parameters @xcite @xmath91 we found that a better fit was attained by dropping the simplifying assumption that @xmath16 is constant across clusters . in fact , we were able to account for all dwarf galaxies smaller than 12 kpc in @xcite by allowing @xmath16 to scale down to : @xmath92 the three schemes we attempted to prescribe for the scale variation of @xmath16 for clusters are as follows : 1 .
@xmath93 , where @xmath89 is the total icm gas mass integrated to @xmath54 , 2 .
@xmath94 , 3 .
@xmath95 .
it is clear that case ( ii ) is the limit of case ( i ) taking @xmath96 and also that case ( iii ) is the limit of case ( i ) taking @xmath97 . by plotting @xmath90 against @xmath89 and
then applying equation ( [ mstgmass ] ) , we were able to constrain the parameter , @xmath16 , using a nonlinear least - squares fitting routine including estimated errors .
the fits for each of the three above schemes are shown in figure [ dynamicmass ] , with the results as follows : 1 .
@xmath98 , 2 .
@xmath99 , 3 .
@xmath100 .
the quality of each of these prescriptions for @xmath16 is seen in their respective plots of figure [ dynamicmass ] where the mstg mass is plotted against the icm gas mass .
clearly , the plot corresponding to case ( i ) is the best scheme since the slope of unity best describes the bulk properties of the clusters .
indeed , the overall least sum of squares best fit is the scheme corresponding to case ( i ) with the other two acting as limiting cases .
meanwhile , both the mond `` convergent mass '' and the newtonian dynamic mass show a discrepancy with the icm gas mass @xcite . in the galaxy rotation curves , we scaled @xmath18 down from equation ( [ galaxyparameters ] ) by a factor of 2 to account for the dwarf galaxies @xcite .
for all but the smallest galaxy clusters where the mass of the icm dominates over the mass of the individual galaxies , scaling @xmath18 up from equation ( [ galaxyparameters ] ) by a factor of 10 leads to satisfactory fits , whereas for the smallest galaxy clusters where the mass of the icm is of the order of galactic masses with @xmath101 scaling @xmath18 to @xmath102 leads to satisfactory fits .
thus our prescription for the running of newton s constant ( rg flow ) for galaxy clusters is fully constrained : @xmath103 @xmath104
we now turn to the individual mass profiles for each cluster .
since the mstg prediction for the mass profile has no free parameters upon enforcing equations ( [ clusterparameters ] ) , we need not perform any nonlinear least - squares fitting for the individual mass profiles as shown in figure [ clusters ] .
the coma cluster is shown first ( with errorbands ) ; and the remaining 105 clusters follow as thumbnails ( without errorbands for clarity ) .
the @xmath1-model is not always a good fit to the data , however , and so the mass profiles shown in figure [ clusters ] have limited reliability and the results of table [ clusterproperties ] show errors ranging from a few percent to in some cases 100% .
@xcite show that some clusters observed with rosat and asca are expressed better by the double @xmath1-model with a second core ; while @xcite show that clusters observed with xmm - newton are well described by an nfw - type profile @xcite showing strong agreement with cdm simulations . while these more sophisticated mass profile models have their advantages , the simplest isothermal isotropic @xmath1-model based upon hydrostatic equilibrium has the fewest parameters and we have used it to show that it is possible and meaningful to fit the x - ray galaxy cluster data without the need for exotic dark matter .
there are also controversial measurements concerning the presence of temperature gradients in galaxy clusters @xcite .
xmm - newton is expected to probe the isothermal properties of the galaxy clusters .
early results such as the @xcite study of the xmm - newton observation of the coma galaxy cluster and the @xcite study of the xmm - newton radial temperature profiles support the isothermal assumption deviating only near the cluster center where the observed x - ray temperatures decrease . in those cases where the uncertainty in the @xmath1-model parameters are sufficiently small ( and therefore the @xmath1-model provides a suitable fit )
, the mstg mass profile traces the icm gas mass profile down to sufficiently small radii and only deviates where the assumption that the icm gas mass is isothermal is no longer valid . at sufficiently small radii ,
the observed x - ray temperatures decrease and thus the mstg prediction based on an isothermal sphere is too large , confirmed by figure [ clusters ] .
a gravity theory consisting of a metric - skew - tensor action ( mstg ) that leads to the modified acceleration law @xcite can be fitted to a large class of x - ray galaxy cluster mass profiles .
the same acceleration law can also be applied to fit a large class of galaxy rotation curves as in @xcite which achieved one parameter fits for the case of the photometric observations with a single parameter @xmath105 ; and also parametric fits which required no simplifying assumption on the constancy of ( m / l ) throughout the galaxy . whereas mond provides a good phenomenological fit to galaxy rotation curves ,
only mstg fits both galaxy rotation curves and the x - ray galaxy cluster mass profiles without the need of introducing a non - baryonic dark matter component .
in addition , whereas mond necessitates the asymptotic `` flat rotation '' velocity out to infinity , mstg only gives the appearance of a `` flat rotation '' velocity within the confines of the galaxy and returns to the familiar newton - kepler form at large distances . using the sloan digital sky survey ( sdss ) ,
@xcite have studied the velocities of satellites orbiting isolated galaxies .
they detected approximately 3000 satellites , and they found that the line - of - sight velocity dispersion of satellites declines with distance to the primary .
the velocity was observed to decline to a distance of @xmath106 kpc for the available data .
this result contradicts the constant velocity prediction equation of mond , but is consistent with the mstg prediction .
the newest results from xmm - newton reveal that the mass profiles of x - ray clusters show a steady rise out to the limit of x - ray observations @xcite and do not show the convergent behavior of mond isothermal spheres .
however , the xmm - newton results are consistent with the mstg isothermal sphere whose densities fall off as @xmath81 without unseen non - baryonic dark matter
. further results from xmm - newton may reveal that the temperature profiles of x - ray clusters are not everywhere isothermal ; and observed gradients may add a greater understanding of the physics of x - ray clusters of galaxies and the law of gravitation that best describes these largest of virialized objects .
this work was supported by the natural sciences and engineering research council of canada .
we thank thomas h.reiprich for supplying the data for our galaxy cluster sample via http://www.astro.virginia.edu/~thr4f/act/gcs/[electronic table ] .
aguirre , a. , schaye , j. & quataert , e. 2001 , apj , 561 , 550 arnaud , m. _ et al . _ 2001 ,
a&a , 365 , l67 arnaud , m. , neumann d.m . , aghanim , n. , gastaud , r. , majerowicz , s. & hughes , j.p .
2001 , a&a , 365 , l80 brownstein , j.r . & moffat , j.w .
2005 , apj , submitted ( http://arxiv.org/abs/astro-ph/0506370[astro-ph/0506370 ] ) cavaliere , a.l . &
fusco - femiano , r. 1976 , a&a , 49 , 137 chandrasekhar , s. 1960 , _ principles of stellar dynamics _ , dover , new york eidelman , s. _ et al . _ 2004 ,
b592 , 1 , 2005 update ( http://pdg.lbl.gov ) king , i.r .
1966 , aj , 71 , 64 milgrom , m. 1983 , apj , 270 , 365 milgrom , m. 1983 , apj , 287 , 571 moffat , j.w .
2005 , j. cosmology astropart .
phys . , 05 , 003 moffat , j.w .
2005 , preprint ( http://arxiv.org/abs/gr-qc/0506021[gr-qc/0506021 ] ) navarro , j.f . ,
frenk , c.s .
, & white , s.d.m .
1997 , apj , 490 , 493 ota , n. & mitsuda , k. 2002 , apj , 567 , l23 pointecouteau , e. , arnaud , m. & pratt , g.w .
2005 , a&a , 435 , 1 pointecouteau , e. & silk , j. 2005 , mnras , submitted ( http://arxiv.org/abs/astro-ph/0505017[astro-ph/0505017 ] ) prada , f. _ et al . _
2003 , apj , 598 , 260 reiprich , t.h .
2001 , ph.d.dissertation , _ cosmological implications and physical properties of an x - ray flux - limited sample of galaxy clusters _ , ludwig - maximilians - universitt mnchen reiprich , t.h . & bhringer , h. 2002 , apj , 567 , 716 reuter , m. & weyer , h. 2004 , j. cosmology astropart .
, 12 , 001 reuter , m. & weyer , h. 2004 , phys .
d69 , 104022 sanders , r.h .
2003 , mnras , 342 , 901 sanders , r.h . & mcgaugh , s.s .
2002 , ara&a40 , 263 the l.s . &
white s , d.m .
1988 , aj , 95 , 1642 zwicky , f. 1933 , helv.phys.acta . , 6 , 110 llll + column ( 1)&galaxy cluster name ( truncated to 8 characters)&column ( 6)&radius where @xmath107 + column ( 2)&x - ray temperature&column ( 7)&icm gas mass integrated to @xmath54 + column ( 3)&icm central mass density&column ( 8)&newtonian dynamic mass integrated to @xmath54 + column ( 4)&@xmath1-model @xmath1 parameter&column ( 9)&mstg dynamic mass integrated to @xmath54 + column ( 5)&@xmath1-model core radius parameter&column ( 10)&convergent mond dynamic mass | we apply the modified acceleration law obtained from einstein gravity coupled to a massive skew symmetric field , @xmath0 , to the problem of explaining x - ray galaxy cluster masses without exotic dark matter . utilizing x - ray observations to fit the gas mass profile and temperature profile of the hot intracluster medium ( icm ) with king `` @xmath1-models '' ,
we show that the dynamical masses of the galaxy clusters resulting from our modified acceleration law fit the cluster gas masses for our sample of 106 clusters without the need of introducing a non - baryonic dark matter component .
we are further able to show for our sample of 106 clusters that the distribution of gas in the icm as a function of radial distance is well fit by the dynamical mass distribution arising from our modified acceleration law without any additional dark matter component . in previous work
, we applied this theory to galaxy rotation curves and demonstrated good fits to our sample of 101 lsb , hsb and dwarf galaxies including 58 galaxies that were fit photometrically with the single parameter @xmath2 .
the results there were qualitatively similar to those obtained using milgrom s phenomenological mond model , although the determined galaxy masses were quantitatively different and mond does not show a return to keplerian behavior at extragalactic distances .
the results here are compared to those obtained using milgrom s phenomenological mond model which does not fit the x - ray galaxy cluster masses unless an auxiliary dark matter component is included .
[ firstpage ] dark matter galaxies : clusters : general galaxies : kinematics and dynamics gravitation x - rays : galaxies : clusters |
despite many efforts devoted , the origin of hydrodynamic turbulence is still poorly understood , because there is a significant mismatch between the predictions of linear theory and the experimental data ( see e.g. @xcite ) .
the mismatch between theoretical results and observed data is also found in astrophysical contexts , where the accretion flow with keplerian angular momentum profile is a common subject . the hot ionized keplerian disc flow , which is linearly stable otherwise , in presence of magnetic coupling is shown to exhibit magneto - rotational - instability ( mri ) @xcite based on a local analysis .
then the unstable modes were thought to lead to nonlinearity and then turbulence in the flow .
however , it was also shown that the criterion drawn from the local dispersion relation is inadequate and may be misleading @xcite .
then the question arises that how could a keplerian accretion disc flow having a very low molecular viscosity generate turbulence , which could successively govern diffusive viscosity to support transfer of mass inwards and angular momentum outwards , in absence of any unstable mode of hydrodynamic origin ? on the other hand , the existence of large transient growth of perturbation in plane couette flow with and without the coriolis force were demonstrated @xcite .
it was argued that the large transient growth plausibly could lead to subcritical transition to turbulence in laboratory experiments and in astrophysical flows which in turn could lead to turbulent viscosity @xcite .
note , however , that the shear flows with and without rotation are physically quite different .
the threedimensional growth factors and then plausible turbulent viscosity in presence of the coriolis force were shown to be significant in keplerian flow only if the elliptical vortices present in the background flow @xcite . the coriolis effect , which absorbs the pressure fluctuation , is the main culprit to kill any growth of energy in the background of linear shear .
moreover , very importantly , the feedback of inherent instability and then sustenance of turbulence based on large transient growth is questionable , in lieu of linear instabilities .
therefore , the problem of hydrodynamic transition to turbulence in rotating shear flow , particularly when angular momentum increases and angular velocity decreases as functions of radial coordinate , e.g. in the astrophysical accretion disc , remains unsolved . while , bech & anderson @xcite found turbulence persisting in numerical simulations of subcritical rotating flows for large enough reynolds numbers , in general the problem of hydrodynamic transition to turbulence in rotating shear flows , particularly the astrophysical ones , has not been paid enough attention in the literature so far .
decades back craik @xcite studied the impact of rotation ( including the keplerian one ) and hence the coriolis effects to the flow with elliptical streamlines . later on , salhi @xcite examined rotation and buoyancy effects on homogeneous shear flows and found similarities between rotation and stratification effects .
in fact , much earlier bradshaw @xcite discussed the formal algebraic analogy between the parameters describing streamline curvature and buoyancy in a turbulent flow .
very recently , the stability problem of unbounded rotating shear flows has been again studied in presence of uniform vertical density stratification @xcite which has important astrophysical applications e.g. in geometrically thin radially stratified accretion discs .
note that indeed johnson and gammie @xcite investigated nonaxisymmetric radially stratified linearized accretion discs in a shearing sheet approximation neglecting vertical structure .
then they studied the time evolution of a plane wave perturbation co - moving with the shear flow . the shearing flow with rotation and stratification
has many other applications , including that in geophysics , as described in detail by salhi & cambon @xcite .
interestingly , same / similar physics is borrowed to understand rotating shear flows in different communities with the use of different terminologies . while engineering community compute rotational richardson number " introduced by bradshaw @xcite to understand linear stability of a flow which is proportional to the square of epicyclic frequency used by the astrophysicists and rayleigh number by the fluid dynamicists in general for the same purpose .
while positive richardson number and rayleigh number correspond to linear stability in engineering and fluid physics respectively , @xmath0 ( real value of epicyclic frequency ) is the equivalent condition in astrophysics with angular velocity @xmath1 ( will be discussed in detail in subsequent sections ) .
we show that rotating flows exhibiting pseudo - eigenmodes , in absence of any unstable eigenmode itself , quickly reach a growth rate of the order of the logarithmic norm under a very small perturbation .
this plausibly triggers turbulence .
note that the logarithmic norm , as discussed later , is an upper bound of such growth rates .
this establishes that the intrinsic perturbation in the system is enough to sustain the turbulence which is a manifestation of the transient local instability .
thus , the turbulence of rotating shear flows and the viscosity of hydrodynamic origin in accretion discs is shown to be an inherent property of these systems . in the next section
we describe hydrodynamic equations in shear flows , pseudo - eigenmodes , logarithmic norm and corresponding transient instability in presence of the coriolis effects .
subsequently , in 3 we discuss solutions .
finally , in 4 we summarize with conclusions .
let us focus on a local analysis considering a small patch of rotating shear flow whose unperturbed velocity profile @xmath2 corresponds to the background shear and angular velocity @xmath1 , as described earlier @xcite , shown schematically in fig .
[ shear ] . following earlier work @xcite all the variables
are written in the dimensionless form .
the figure describes our choice of local cartesian coordinates : @xmath3 is along the radial direction , @xmath4 is along the azimuthal or streamwise direction , and @xmath5 is along the vertical direction .
the unperturbed flow has a velocity in the @xmath4-direction and a gradient of velocity i.e. shear along the @xmath3-direction .
the coriolis force arised due to rotation of the patch is described by an angular velocity vector @xmath6 along the @xmath5-direction . for convenience ,
we write down the corresponding equations for evolution of perturbation in terms of the set of velocity and corresponding vorticity @xmath7 and hence the orr - sommerfeld and squire equations in the presence of coriolis acceleration are given by @xmath8 @xmath9 where @xmath10 is the @xmath3-component of velocity and @xmath11 is the corresponding component of vorticity , @xmath12 for a keplerian flow and @xmath13 for a constant angular momentum flow , and @xmath14 is the reynolds number .
note that in presence of linear background shear , what is of present interest , @xmath15 and hence the term proportional to @xmath16 vanishes . for detailed description ,
see mukhopadhyay et al .
we further reduce the set of equations in the form of an eigenvalue equation @xmath17 where @xmath18,\\ \zeta(x , y , z , t)=\hat{\zeta}(x , t ) \exp[i(k_yy+k_zz ) ] , \label{sol5}\end{aligned}\ ] ] and @xmath19 , is a differential operator . note that
because of translation invariance of the unperturbed flow in the @xmath4 and @xmath5 directions , we decompose the perturbations in terms of fourier modes . the solution of eqn .
( [ sol6 ] ) , describing the evolution of perturbation , can be written in terms of an eigenfunction expansion @xmath20 where the complex eigenvalue of the problem @xmath21 .
therefore , for the @xmath22th mode , eqn .
( [ sol6 ] ) reduces to @xmath23 to obtain the set of eigenvalues and eigenvectors , we convert the differential operator @xmath24 into an @xmath25 matrix in a finite - difference representation and then compute the eigenvalues and eigenvectors of the matrix .
the physical parameters involved in the problem ( mainly @xmath14 and the components of perturbation wave - vector @xmath26 described below ) determine the required order @xmath27 of the matrix for adequate accuracy .
it is well known ( see e.g. @xcite ) that all the eigenmodes are stable , i.e. @xmath28 for all @xmath22 , when @xmath0 .
however , the corresponding pseudo - eigenmodes may be unstable .
following previous work discussing nonrotating shear flows @xcite , let us try to understand it .
if the linearized fluid system is driven by a signal of the form @xmath29 , then eqn .
( [ sol6 ] ) becomes @xmath30 it can be verified that the response will be of the form @xmath31 where @xmath32 and @xmath33 is the identity matrix .
the operator @xmath34 transforms the input @xmath35 to the linearized flow at frequency @xmath36 into the corresponding output @xmath37 .
the maximum amplification rate that occurs in the process can be written as @xmath38 where @xmath39 denotes a norm of the respective quantities , in the present work @xmath40norm @xcite .
now an eigenvalue of @xmath24 is @xmath36 if @xmath41 where @xmath37 is the corresponding eigenvector . on the other hand , using eqns .
( [ amp ] ) and ( [ eigen ] ) @xmath36 can be identified as an eigenvalue of @xmath24 if the perturbation with frequency @xmath36 brings an unbounded amplification so that @xmath42 generalizing the above definition of an eigenvalue , we can define @xmath36 as an @xmath43-pseudo - eigenvalue @xcite of @xmath24 for any @xmath44 , if @xmath45 when @xmath37 is the corresponding @xmath43-pseudo - eigenvector / mode such that @xmath46 naturally as @xmath47 , @xmath37 and @xmath36 tend to be the pure eigenmode and eigenvalue respectively .
the set of @xmath43-pseudo - eigenvalues of an operator is called the @xmath43-pseudo - eigenspectrum .
therefore , we can classify the amplification in three regimes .
first , if @xmath14 is equal or greater than a certain critical value @xmath48 ( @xmath49 ) , and @xmath24 has at least one _ eigenvalue _ in the upper half - plane of the argand diagram , then @xmath50 and @xmath51 ; the flow in this regime is linearly unstable ( i.e. the system has at least one exponential growing mode ) .
second , if @xmath14 is smaller than a certain value , @xmath52 ( @xmath53 ) , and all the _ eigenvalues _ lie in the lower half - plane of the argand diagram , then @xmath54 and @xmath55 ; this implies no amplification ( i.e. there is no exponential growing mode ) . finally ,
if @xmath56 and @xmath24 has at least one pseudo - eigenvalue in the upper half - plane ( although all the eigenvalues still lie in the lower half - plane ) , then @xmath57 and @xmath58 ; this is the regime of linearly stable flow with the existence of transient growth ( i.e. the system has at least one growing pseudo - eigenmode ) . now
a very important related quantity is the logarithmic norm which is the upper bound of the possible transient growth rate .
the logarithmic norm , @xmath59 , of a square matrix operator @xmath60 using a matrix @xmath61-norm is defined @xcite as @xmath62 for a natural number @xmath63 , where @xmath64 is the time interval , @xmath65 is the set of real numbers . for @xmath66
, it is readily shown that @xmath67\right ) , \label{mu2 } \end{aligned}\ ] ] where @xmath68 denotes an eigenvalue of the symmetric operand matrix and @xmath69 indicates real part .
the growth of the solution @xmath70 of eqn .
( [ sol7 ] ) over a non - zero time interval @xmath64 can be upper bounded @xcite as follows : @xmath71 and the limiting maximum growth rate then can be written as @xmath72 from which one can see that the logarithmic norm of @xmath73 determines the upper bound of the transient growth of the system in eqn .
( [ sol7 ] ) .
hence , for a suitably small perturbation if we obtain a pseudo - eigenvalue which is close to the logarithmic norm and is in the upper half plane of the argand diagram , then the system will exhibit sustainable transient instability .
the non - negative real parts in the pseudo - eigenspectrum of @xmath73 , i.e. the values in the upper half plane of the argand diagram for the pseudo - eigenspectrum of @xmath74 , drive the transient growth .
the pseudo - eigenvalue of @xmath75 can be defined , equivalently to eqn .
( [ epeigen ] ) , as @xcite @xmath76 using the triangular inequality property of the logarithmic norm , we can write @xmath77 \le \mu_2(-i { \mathcal l } + e ) \le \mu_2(-i { \mathcal l } ) + \epsilon . \label{lognormeps}\end{aligned}\ ] ] the above inequality leads to the computation of a minimum @xmath43 ( namely @xmath78 ) for which the equality holds .
if @xmath78 is small ( compared to the matrix @xmath79-norm of the operator , i.e. , @xmath80 ) such that @xmath81 > 0 $ ] , i.e. , there are pseudo - eigenvalues in the upper half plane of the argand diagram , then we can conclude that the rotational shear flow system with the coriolis effect exhibits transient growth sustained by small intrinsic perturbations .
we solve the eigenvalue equation ( [ sol66 ] ) with no - slip boundary conditions @xmath82 in table 1 we enlist the optimum sets of parameters governing the perturbation that induce maximum growth for various values of @xmath83 .
we compute the maximum transient growth @xmath84 induced by the perturbation and the pseudo - eigenvalues determining the lower bound of the transient growth of energy : @xmath85 , where @xmath86 indicates imaginary part .
interestingly , while for @xmath13 the optimum perturbation giving rise to the maximum growth is purely vertical , for @xmath87 it is purely twodimensional .
for @xmath88 it is threedimensional with @xmath89 and @xmath90 both nonzero .
figure [ fig1 ] shows the optimum pseudo - eigenspectra for various values of @xmath83 . naturally for @xmath13 the spectrum is marginally unstable and the pseudo - eigenspectra for all values of @xmath43 extend to the upper half plane in the argand diagram .
however , as @xmath83 decreases , particularly from @xmath91 to @xmath92 , the pseudo - eigenvalues for particular values of @xmath43 , e.g. @xmath93 and @xmath94 , decrease , which in turn decrease the rate of maximum growth as shown in table 1 . as the maximum growth for @xmath13 is due to the pure vertical perturbation
, we show in fig .
[ fig2 ] the pseudo - eigenspectra due to the pure vertical perturbation , giving rise to the maximum possible growth ( maximal pure vertical perturbation ) , for @xmath0 as well .
corresponding parameters are given in table 2 .
interestingly , the spectra and thus growth for a particular @xmath83 appear quite different between that for the optimum and vertical perturbation ( particularly for smaller values of @xmath83 ) .
this is due to the coriolis effect generating epicyclic oscillations appeared in the later case , which is responsible for hindering growing pseudo - eigenmodes .
note that the energy growth due to pure vertical perturbation is bounded by the dynamics of the system and is independent of @xmath14 . in fig .
[ fig4 ] we compare pseudo - eigenspectra of the flow with @xmath87 and @xmath95 for two different values of @xmath14 and find them very similar , reflecting their independence of @xmath14 under pure vertical perturbation .
table 3 shows the minimum perturbation ( @xmath78 ) for which the logarithmic norm rate of growth is approached .
it may be noted that @xmath96 . also , at @xmath13 the flow is most susceptible to transient growth since the required perturbation is the least .
a decrease in @xmath83 reduces the susceptibility to transient growth as the magnitude of the required perturbation increases .
however , @xmath78 is still quite small even for @xmath87 , assuring local instability and sustained turbulence .
while the existence of unstable pseudo - eigenmodes favors instability and then plausible turbulence under finite amplitude perturbation , the question remained whether the local instability is sustained or not .
the present analysis argues that the rotating shear flows with the coriolis force not only exhibit unstable pseudo - eigenmodes under suitable small perturbations , but also the growth rate in the pseudo - eigenmodes approaches a strictly positive logarithmic norm and thus assures feedback of instability as an intrinsic property of the flows .
therefore , a very small amplitude of perturbation is able to trigger instability and then plausible turbulence in the system . in lieu of linear instabilities ( of hydrodynamic and hydromagnetic origin )
unstable pseudo - eigenmodes repopulating the growing disturbance assure _ sustained _ turbulence .
this argues , for the first time , in favor of _ sustained _ hydrodynamic instability and then turbulence in a rotating couette like flow , when angular momentum increases and angular velocity decreases with radius .
this is practically the flow of many astrophysical systems , e.g. accretion disc around compact objects , some active galactic nuclei , star - forming systems .
this is possible to conceive because of the non - self adjoint nature of underlying operator , giving rise to the non - normal eigenmodes and hence unstable pseudo - eigenspectrum and positive logarithmic norms .
note that molecular viscosity is too small to explain the observed accretion efficiencies in the systems mentioned above .
therefore , any viscosity is expected to arise due to turbulence in the flow .
hence , in absence of any linear instability in the keplerian accretion flows , existence of small perturbations triggering local transient growth rates of the order of logarithmic norm , which is strictly positive for the associated spatial operator matrix , resolve the question of instability and corresponding feedback process which replenish the growth to an ultimate turbulence state .
this is a new alternate picture of turbulence arousal in astrophysical flows explaining the transport of mass inward removing angular momentum outward , particularly when transport due to the nonlinearly developed mri is expected to be very small @xcite in experimental setups with small magnetic prandtl number .
0.2 cm table 1 parameters for optimal / maximal perturbation with @xmath97 0.2 cm table 2 parameters for maximal pure vertical perturbation with @xmath97 table 3 minimum perturbation for logarithmic norm growth rate at @xmath97
the authors would like to thank the referee for his / her comments which helped to improve the presentation of the paper .
two of the authors ( bm , rm ) were supported in part by dst grant sr / s2hep12/2007 funded by government of india . another author ( sr ) was partially supported by the iisc mathematics initiative ( imi ) programme . | rotating shear flows , when angular momentum increases and angular velocity decreases as functions of radiation coordinate , are hydrodynamically stable under linear perturbation .
the keplerian flow is an example of such systems which appears in astrophysical context .
although decaying eigenmodes exhibit large transient energy growth of perturbation which could govern nonlinearity into the system , the feedback of inherent instability to generate turbulence seems questionable .
we show that such systems exhibiting growing pseudo - eigenmodes easily reach an upper bound of growth rate in terms of the logarithmic norm of the involved nonnormal operators , thus exhibiting feedback of inherent instability .
this supports the existence of turbulence of hydrodynamic origin in the keplerian accretion disc in astrophysics .
hence , this enlightens the mismatch between the linear theory and experimental / observed data and helps in resolving the outstanding question of origin of turbulence therein . |
soudan-2 is a fine - grained iron tracking calorimeter located 2100 mwe underground in soudan , minnesota , usa . the experiment has been taking data since 1989 when the detector was one quarter of its full size ; the construction was completed at the end of 1993 .
the experiment continues to take data with 90% live time , having accumulated 3.6 kty up to the date of this conference .
the first result on the ratio of ratios of the atmospheric neutrinos , @xmath2_{data } } \over{[(\nu_\mu+\bar\nu_\mu)/(\nu_e+\bar\nu_e)]_{mc } } } \label{eq : r}\ ] ] was published earlier this year.@xcite
the two main components of the soudan-2 apparatus are the central detector and the active shield . the _ central detector _
consists of 224 @xmath3 m calorimeter modules weighing 4.3 tons each .
the bulk of the mass consists of 1.6-mm - thick corrugated steel sheets which are interleaved with drift tubes .
each half of a 1.5-cm - diam .
, 1-m - long , resistive drift tube is read out by a crossed pair of a vertical anode wire and a horizontal cathode pad that determine two coordinates of a tube crossing ; the third coordinate is deduced from the drift time . readout from the 3.4 million half - tubes is accomplished using 136,000 anode and cathode channels that are multiplexed before the signals are digitized .
the _ active shield _ covers the walls of the soudan-2 cavern enclosing the central detector as hermetically as possible . only the calorimeter s support columns must pass through the shield , adding up to about 5 m@xmath4 , out of the total shield area of 1700 m@xmath4 .
the shield consists of 1529 modules with a typical size of 19 cm @xmath5 7 m , each containing a double layer of proportional tubes with a digital readout .
a coincidence of two adjacent layers is required for a hit in the off - line analysis .
detailed descriptions have been published of the design and performance of soudan-2 calorimeter modules @xcite , and of the active shield system @xcite .
the contained events ( cev ) in an underground detector are expected to be atmospheric neutrino interactions , where the neutrinos originate in decays of mesons and muons in cosmic ray showers in the atmosphere .
it is these interactions that we model in our monte carlo simulation and compare with data .
we use the bartol atmospheric neutrino flux @xcite , and every mc event is overlayed onto a random pulser event to acquire a realistic noise pattern . in the following , we used simulated events corresponding to an 18.3 kty exposure of soudan-2 ; the dependence of the neutrino flux on the solar cycle was taken into account . during the period april 1989 through december 1996 , we analyzed 75 million triggers , most of which are initiated by throughgoing muons or noise " ( both electronic noise and triggers due to local radioactivity in the cavern ) .
75% of the monte carlo events pass the hardware trigger requirement .
all triggers are processed by a two - stage software filter .
the software performs a number of data quality checks , and rejects events that conform to one of the identified noise patterns . to check for event containment
, the software requires that _ ( i ) _ no part of the event is within 20 cm of the detector surface , and that _ ( ii ) _ no track is located and oriented in such a way that it could enter the calorimeter undetected ( e.g. through a crack between modules ) .
the number of events satisfying the software requirements is 15,000 per kty .
about 50% of monte carlo events that satisfy the trigger conditions pass the software filter .
the last step in the data reduction chain is a two - stage physicist scan .
the resulting cev sample contains typically 500 events per kty , representing an overall data reduction of @xmath6 .
the monte carlo events are injected into the data stream before the scanning stage so that physicists scan data and mc simultaneously without knowing which is which . of the remaining monte
carlo events , 70% pass the scan requirements , indicating an overall neutrino cev detection efficiency of about 26% . in order to measure the @xmath7 flavor composition in as unbiased a way as possible
, it is customary to use a ratio of ratios " .
we express @xmath8 defined in eq .
[ eq : r ] using our number of tracks , @xmath9 , as a measure of @xmath10 events , and number of showers , @xmath11 , as a measure of @xmath12 events : @xmath13 before the background correction , we obtain @xmath14 using the raw gold sample counts . using the corrected numbers of neutrino events in our gold sample from table 2 , we obtain @xmath15 where the first error is statistical , and the second is systematic .
our value of @xmath8 is consistent with results of underground water detectors reviewed in ref .
the indication is that there are too few single tracks , i.e. too few @xmath10 and @xmath16 events in the underground neutrino flux , usually explained as due to neutrino oscillations . finally , we present lepton energy distributions for our track sample in fig .
[ fig : energy_t ] and for our shower sample in fig .
[ fig : energy_s ] . in the upper portion of the figures ,
the crosses represent our gold data before any correction , and the solid histogram shows the rock background .
the dashed histograms are the monte carlo distributions . in the lower portion of the figures
we show our corrected data distributions .
we note that the background is confined not only to smaller penetration depths , but also to smaller energies than the neutrino event distributions
. 9 w.w.m .
et al . _ ,
b391 ( 1997 ) 491 .
et al . _ ,
instr . meth .
a376 ( 1996 ) 36 ; _ ibid _ a381 ( 1996 ) 385
et al . _ , nucl .
instr . meth
. a276 ( 1989 ) 371 .
g. barr , t.k .
gaisser , t. stanev , phys .
d39 ( 1989 ) 3532 .
e. kearns , these proceedings . | a new measurement of the atmospheric @xmath0 ratio - of - ratios , @xmath1 , has been obtained using a 3.2-kty exposure of the soudan-2 underground detector .
this measurement , based upon neutrino reactions in an iron tracking calorimeter of honeycomb - lattice geometry , is in agreement with the anomalously low value reported by the underground water detectors . |
the central parsec of the milky way contains a large population of massive o / wr and b stars both as part of a stellar disc @xcite and in a cusp extending towards the galactic center ( gc ) black hole , sgr a * ( the s cluster , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the orbits of these stars around sgr a * , particularly the star s2/s0 - 2 , measure the black hole mass to be @xmath2 ( e.g. , * ? ? ?
the detection and timing of pulsars in similar orbits around sgr a * would provide precise measurements of the black hole mass , and potentially detect several other general relativistic effects @xcite . based on the large number of massive stars and a significant enhancement in the x - ray binary density @xcite , it is widely expected that the central parsec hosts a large neutron star population .
@xcite predicted that @xmath3 pulsars should have a semi - major axis @xmath4pc from sgr a*. it has long been thought that the strong interstellar scattering screen towards the gc , known for decades to blur vlbi images of sgr a * ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , prevented the efficient detection of pulsars .
if the scattering material is close to the gc , it would strongly broaden pulse profiles , rendering pulsars undetectable at frequencies @xmath5 ghz @xcite .
high frequency ( @xmath6 ghz ) pulsar searches have been performed over a range of beam sizes and sensitivities @xcite .
none of these searches have discovered pulsars in the central parsec , and the deepest search at high frequency ( @xmath7 ) already would be sensitive to a significant fraction of known pulsars @xcite .
recently , _
swift _ detected an @xmath8-ray outburst from the gc @xcite .
_ nustar _ discovered pulsations with a period of @xmath9 s @xcite , and _
chandra _ localized the source to a projected separation of only @xmath10 pc from sgr a * @xcite .
radio pulsations were subsequently detected , allowing for measurements of the dispersion and rotation measures @xcite .
the dispersion measure was the highest ever recorded for a pulsar and placed it in the central parsec .
using vlbi observations , @xcite found the angular broadening for this new highly magnetized neutron star ( magnetar ) , sgr j1745 - 29 , was in excellent agreement with that of sgr a * , indicating that both sources are behind the same scattering screen , which implies that the screen is uniform on @xmath11 pc scales .
the temporal broadening , however , was much less than previously expected @xcite , and similar to pulsars discovered on larger scales ( @xmath12 ) .
the angular broadening is also comparable to observations of masers in the gc from @xmath13
@xmath14 pc @xcite .
these results suggest that past failures to discover pulsars at low frequencies are not simply due to strong interstellar scattering . in light of these observations ,
we assess the central parsec pulsar population .
we assume that the scattering screen is far from the gc , and uniformly apply the temporal broadening measurements of @xcite to the central parsec .
we show the impact of these results on previous observational constraints , and argue that the failure to detect ordinary pulsars in past surveys constitutes a `` missing pulsar problem '' ( [ sec : observ - constr ] ) .
in contrast , the detection of even a single magnetar is unexpected , and implies a high magnetar formation efficiency .
this efficiency is sufficient to explain the final outcome of all of the massive o / b stars ( [ sec : effic - magn - form ] ) .
millisecond pulsars ( msps ) are not as well constrained , but if present should be detectable in future high - frequency pulsar surveys ( [ sec : galact - centre - mill ] ) .
we discuss possible solutions to the missing pulsar problem and implications of these results in [ sec : discussion ] . a strongly inhomogeneous scattering medium , while plausible , is unsupported by current data .
an intrinsic pulsar deficit is a more likely explanation for the lack of detections .
the effect of neutron star kicks on isolated pulsars born in the central parsec .
the cumulative fraction of pulsars born in the disk of young stars is shown as a function of projected distance at birth ( dotted , * ? ? ?
* ) , and at typical ages of magnetars ( dashed ) and ordinary pulsars ( solid ) . in both cases
the retention fraction of both populations in the central parsec is @xmath15 .
nearly all ( @xmath16 ) of the magnetars remain within the _ swift _ field of view ( @xmath17 pc ) .
the expected cumulative fraction of pulsars within @xmath18 pc depends on the assumed orientation ( as shown by the shaded region ) and inner edge of the stellar disc . ]
observational constraints on the gc pulsar population .
the dots show the luminosity vs. period of known pulsars @xcite .
the lines show the sensitivity limits of past surveys and are cut off at @xmath19 as in @xcite , where @xmath20 is the previously assumed ( dashed , * ? ? ?
* ) and measured ( solid , * ? ? ?
* ) temporal scattering towards sgr j1745 - 29 .
the black lines are from the @xmath21 ghz survey of @xcite , while the tightest constraints , shown as the red lines , come from the effelsberg @xcite and arecibo @xcite surveys at @xmath22 ghz .
approximately half of known ordinary pulsars could have been detected by past surveys .
the luminosity of sgr j1745 - 29 re - scaled to @xmath23 and @xmath24 ghz ( red and black stars ) is well above the detection threshold of both surveys .
the @xcite survey at @xmath24 ghz should be sensitive to pulsars down to the shortest known periods . ]
pulsars are rotating , magnetized neutron stars formed from the collapse of massive stars ( @xmath25 ) like those in the gc .
pulsars spin down over time , with a characteristic age , @xmath26 where @xmath27 and @xmath28 are the magnetic field strength and rotation period , scaled to typical values of known pulsars ( referred to here as `` young '' or `` ordinary '' pulsars ) .
magnetars with strong surface fields ( @xmath29 g ) spin down much more rapidly and have correspondingly shorter lifetimes ( @xmath30 yr ) . in either case , eventually the spin period becomes so long that pulsar emission turns off ( the `` death line '' ) .
the expected total number of young pulsars in the central parsec , @xmath31 , can be estimated from the number of massive stars and the relative ages of the stars ( @xmath32 myr ) and pulsars @xcite : @xmath33 where @xmath34 , and @xmath35 is a lower limit to the total number of massive stars in the central parsec ( e.g. , * ? ? ?
* spectroscopically identified 177 wr / o / b stars in the central parsec with @xmath36 completeness . ) .
the parameter @xmath37 is the pulsar formation efficiency from massive stars , and @xmath38 is the fraction of pulsars which are retained in the inner parsec after their natal kicks
. pulsars with typical kick velocities , @xmath39 , can escape the central parsec and deplete the pulsar population .
figure [ retention ] shows the expected spatial distribution of neutron stars after @xmath40 yr , assuming that they are kicked from a circular disk of stars similar to that observed in the central parsec @xcite with three - dimensional kick velocities following the double - exponential distribution of @xcite .
orbits were integrated using the package galpy , with a model potential that reproduces the circular velocity of stars from sgr a * through the milky way halo .
approximately half of the isolated pulsars and magnetars escape the inner parsec within their lifetime .
nearly all magnetars remain within the central ten parsecs , within the field of view of the _ swift _ @xmath8-ray telescope .
roughly @xmath41 of the ordinary pulsars can be found outside the central parsec , where they formed , but within @xmath42pc , within the field of view of some previous pulsar surveys .
however , @xcite point out that observations of globular clusters find large pulsar populations despite escape velocities much smaller than the average pulsar kick velocity .
this is presumably due to the larger escape velocity needed to escape binary systems , combined with high binary fractions in dense stellar environments .
similar effects could weaken the impact of neutron star kicks in the gc .
we conservatively adopt @xmath43 in the following , but this is likely a lower limit .
the number of observable pulsars in the gc depends on the total number of pulsars ( eq . [ eq:2 ] ) , as well as the survey parameters , pulsar luminosity and period distributions , and the fraction beamed towards us ( assumed to be @xmath44 throughout ) : @xmath45 where @xmath46 is the detection efficiency of ordinary pulsars in each survey .
past surveys estimated low values of @xmath46 based on the large expected temporal scattering towards the gc .
we re - assess the detection efficiencies in light of the measured temporal broadening to sgr j1745 - 29 @xcite .
figure [ constraints ] shows the luminosities and periods of known pulsars @xcite , as well as the limits from the deepest surveys at @xmath22 @xcite and @xmath24 @xcite ghz .
the previously assumed constraints are shown as dashed lines , while the solid lines use the measured degree of temporal broadening towards the gc .
the @xmath22 ghz results provide the strongest constraint on the ordinary pulsar population , while the @xmath24 ghz survey may be sensitive to bright msps ( [ sec : galact - centre - mill ] ) .
similar results are obtained using the sensitivity of the @xmath47 ghz survey of @xcite .
we estimate @xmath46 as the fraction of known ordinary pulsars ( @xmath48 ms ) with luminosity larger than the detection threshold of each survey .
the detection thresholds are extrapolated to @xmath49 ghz assuming a spectral index of @xmath50 @xcite .
we find consistent results by instead using the observed spectral indices to extrapolate the pulsar fluxes to the frequencies of the surveys .
we roughly account for possible malmquist bias in the pulsar distribution by doing the estimate separately for all and the closest @xmath51 of pulsars .
the result is @xmath52 at @xmath53 ghz and @xmath54 at @xmath21 ghz . for @xmath55 ,
@xmath56 estimated above , and @xmath57
as expected , the deepest survey should have detected @xmath58 ordinary young pulsars .
given an expectation value of @xmath59 , it is statistically improbable ( @xmath60 ) to have no detections .
although the systematic uncertainties in this estimate are larger than the statistical ones , they are likely insufficient to explain the discrepancy ( @xmath61 for the smallest value of @xmath46 at @xmath22 ghz ) .
this is especially true given our conservative choices for the retention fraction and total number of massive stars .
a simple alternative estimate for the expected number of detectable pulsars can be made by scaling the number of observed pulsars ( @xmath62 ) to the fraction of galactic supernovae that occur in the gc ( @xmath63 ) : @xmath64 , consistent with the above .
we conclude that there is a missing pulsar problem in the gc . at @xmath65 , an upper limit to the total number of young pulsars
is @xmath66 .
an alternative view of this constraint is as an upper limit to the pulsar formation efficiency of @xmath67 .
that is , @xmath68 of massive stars in the gc do not form ordinary pulsars .
we note that young pulsars make up a small portion of the overall population , and so there could be many more old and/or faint pulsars as discussed by @xcite .
this does not change our estimates of @xmath69 or @xmath37 .
in contrast to the missing pulsars , the detection of a young magnetar ( @xmath30 yr ) so close to sgr a * implies a large population of highly magnetized neutron stars in the central parsec . the expected number of observable magnetars , @xmath70 can be estimated in the same way as above , but for the much shorter magnetar lifetime :
@xmath71 where the parameters are scaled to the properties of sgr j1745 - 29 and @xmath72 is the detection fraction of magnetars .
we would not have expected to see a magnetar unless magnetar formation is efficient ( @xmath73 ) .
long after a pulsar passes the death line it may accrete mass from a stellar companion , and be rejuvenated in an @xmath8-ray binary ( xrb ) as an msp .
the process of forming msps is greatly enhanced in dense environments by dynamical encounters between neutron stars and binary stars @xcite .
indeed , old , massive star clusters have some of the richest populations of msps and xrbs .
@xcite found that the encounter rate and stellar mass in the gc is similar to the globular cluster terzan 5 , which has a large population of msps .
they estimated that the gc population could be as large as @xmath74 total msps , accounting for its higher pulsar retention fraction .
a large population of msps in the gc is further supported by the overabundance of xrbs found near sgr a * @xcite . to date
this population has remained mostly unconstrained .
_ fermi _ observations are consistent with up to @xmath75 msps within the central @xmath76 pc @xcite .
even the relatively small degree of temporal scattering measured by @xcite precludes the efficient detection of msps at @xmath77 ghz .
however , the @xcite survey at @xmath78ghz could detect the most luminous @xmath79 @xmath80 of known msps .
if the gc msp population were the same as that of terzan 5 , 2 msps should have been detected by the @xcite survey .
while the number of msps may be enhanced in the gc by dynamical encounters , the estimates typically assume a large population of old neutron stars with low magnetic fields .
if all the neutron stars are magnetars rather than ordinary pulsars , significant field decay would be required to allow spin up to millisecond periods .
despite predictions of a large pulsar population in the central parsec and several deep radio surveys , no ordinary pulsars have been detected to date .
given the recent measurements of weak temporal broadening of the pulses of the magnetar sgr j1745 - 29 @xcite , previous surveys should have seen @xmath81 ordinary pulsars in the central parsec ( [ sec : observ - constr ] ) .
we have argued that this discrepancy constitutes a `` missing pulsar problem '' in the galactic center .
the lack of pulsars could be caused by a low formation efficiency , a low detection efficiency , or a low retention fraction of ordinary pulsars .
here we address these possibilities .
in particular we argue that a strong but inhomogeneous scattering medium that lowers the detection efficiency of surveys is not the most natural explanation for the missing pulsars . _ a patchy scattering screen : _ we have assumed a uniform scattering screen across the inner parsec , as shown to fit both the temporal and angular broadening of sgr j1745 - 29 @xcite .
an alternative explanation suggested by @xcite is that the discovery of a rare , short - lived magnetar is evidence for a large population of ordinary pulsars as well .
these pulsars may not have been seen if the small degree of temporal broadening of sgr j1745 - 29 is explained as a lucky line of sight through a much stronger but inhomogeneous scattering screen . however , multiple lines of evidence suggest that the angular and temporal scattering may be roughly uniform on scales of tens of parsecs . @xcite
found excellent agreement in the angular broadening properties of sgr j1745 - 29 and sgr a * , suggesting that the screen has a coherence length @xmath11 pc . on larger scales ,
pulsars detected within @xmath82 pc have temporal scattering consistent with or less than that of sgr j1745 - 29 @xcite .
similar angular broadening is seen in many oh / ir stars at separations @xmath13 @xmath83 @xcite .
furthermore , the extragalactic source g359.87 + 0.18 @xcite has an angular size consistent with sgr a * at @xmath84 ghz , as would be expected for a uniform but distant screen .
since all of the data can be explained by a uniform but distant screen , and in the absence of direct evidence for strong temporal scattering , an intrinsic deficit in the ordinary pulsar population is the more natural interpretation of the lack of detections in surveys to date
. _ detection efficiency biases : _ if the luminosity and period distributions of galactic center pulsars are similar to those detected elsewhere in the galaxy , the detection efficiency from previous surveys should be high ( [ sec : observ - constr ] ) . however
, this distribution is subject to selection effects , and the true distribution could be biased to fainter pulsars .
we estimate that this possible bias could contribute at the factor of a few level by separately considering only relatively nearby pulsars , which is probably insufficient to explain the lack of detections .
a simple empirical estimate based on the fraction of galactic supernovae occurring in the gc also gives consistent results , and is not subject to this bias . _
neutron star kicks : _ we found that @xmath85 of the young pulsars could be expelled from the central parsec from natal kicks ( [ sec : isol - puls - kicks ] ) .
this is likely an upper limit , as indicated by the large populations of pulsars in globular clusters , which have much smaller escape velocities .
neutron star kicks can not explain the missing pulsar problem . _ peculiar star and pulsar formation : _ the detection of even a single magnetar so close to sgr a * is unexpected , and implies order unity efficiency for forming magnetars from massive stars in the galactic center .
this implied high magnetar efficiency is one possible explanation for the missing pulsar problem : massive stars in the central parsec may form short - lived magnetars rather than long - lived ordinary pulsars .
efficient magnetar formation could be explained by an unusual progenitor population .
there is some evidence that the imf in the central parsec is top - heavy @xcite , and several magnetars are associated with massive stellar progenitors ( @xmath86 , * ? ? ? * ; * ? ? ?
* ; * ? ? ?
* compared to @xmath87 typically assumed for neutron stars ) .
stars in the central parsec may also be highly magnetized ( e.g. , s0 - 2/s2 @xcite , see also the hypervelocity stars , @xcite ) .
magnetars could also be formed from amplification of the field of a highly magnetized progenitor during core collapse @xcite .
both a top - heavy imf and strong magnetic fields could be caused by rapid star formation in the dense regions near sgr a * @xcite .
the star formation event leading to the currently observed massive stars in the central parsec was too recent to have formed ordinary pulsars .
another explanation for missing young pulsars is that past star formation events produced fewer ( by @xmath88 ) massive stars ( e.g. , from a different imf ) .
a very recent major merger between sgr a * and another massive black hole could also have removed a significant fraction of the old stars and pulsars from the central parsec . finally , it is possible that the ordinary pulsars are present , but that the pulsar emission is suppressed , e.g. , by interactions of the neutron star magnetosphere with the gc environment over the pulsar lifetime .
whatever the explanation , missing pulsars and efficient magnetar formation constitute evidence that neutron star formation and evolution are substantially modified in the high density regions near a massive black hole . the authors are listed in alphabetical order . j.d . and r.m.o
. contributed equally to the writing and calculations presented in this study .
we thank g. bower , m. kerr , c. law , and a .- m .
madigan for helpful discussions related to this work , and the participants of iau symposium 303 for coining the phrase `` missing pulsar problem . ''
we thank the referee , s. nayakshin , for helpful suggestions , and j. bovy for publicly releasing the software package galpy , which we used in this work . | pulsars orbiting the galactic center black hole , sgr a * , would be potential probes of its mass , distance and spin , and may even be used to test general relativity . despite predictions of large populations of both ordinary and millisecond pulsars in the galactic center
, none have been detected within @xmath0 pc by deep radio surveys .
one explanation has been that hyperstrong temporal scattering prevents pulsar detections , but the recent discovery of radio pulsations from a highly magnetized neutron star ( magnetar ) within @xmath1 pc shows that the temporal scattering is much weaker than predicted .
we argue that an intrinsic deficit in the ordinary pulsar population is the most likely reason for the lack of detections to date : a `` missing pulsar problem '' in the galactic center .
in contrast , we show that the discovery of a single magnetar implies efficient magnetar formation in the region .
if the massive stars in the central parsec form magnetars rather than ordinary pulsars , their short lifetimes could explain the missing pulsars .
efficient magnetar formation could be caused by strongly magnetized progenitors , or could be further evidence of a top - heavy initial mass function . furthermore , current high - frequency surveys should already be able to detect bright millisecond pulsars , given the measured degree of temporal scattering . |
kmgf@xmath0 is a technologically important flouroperovskite . for example
, it is used as a vacuum - ultraviolet - transparent ( vuv - transparent ) material for lenses in optical lithography steppers@xcite and in electro - optical applications@xcite . when doped with lanthanide ions ,
it is a very promising material for scintillators@xcite and radiation dosimeters@xcite .
in addition the physical properties of kmgf@xmath0 may have implications for understanding of the earth s lower mantle@xcite .
kmgf@xmath0 was first synthesized by van arkel@xcite and has a simple cubic perovskite structure at room temperature@xcite .
kmgf@xmath0 demonstrates great stability under high compression and has not been found to undergo any phase transition at any temperature or pressure , suggesting it may be used as an internal x - ray calibrant@xcite .
several experimental studies of the ground state properties of kmgf@xmath0 have been performed .
the elastic constants at ambient pressure have been measured by rosenberg and wigmore @xcite and by reshchikova,@xcite while jones investigated their pressure and temperature dependence@xcite . from the theoretical side , electronic structure calculations for kmgf@xmath0
have been carried out by means of linear combination of atomic orbitals@xcite , including the effects of doping with transition metal impurities in kmgf@xmath0@xcite .
the electronic structures of divalent 3@xmath2 transition metal impurities doped in kmgf@xmath0 have been investigated by the pseudopotential method@xcite , and the properties of vacancies were studied by hartree - fock cluster calculations@xcite . the structural , electronic and optical properties of kmgf@xmath0
were recently investigated by the full - potential linear augmented plane wave ( fp - lapw ) method@xcite .
the present work is a combined theoretical and experimental study of the ground state and high - pressure properties of kmgf@xmath0 .
we present the equation of state resulting from high - pressure diamond - anvil cell experiments on kmgf@xmath0 up to 40 gpa .
we also present the equation of state , the elastic constants and the electronic structure from theoretical calculations using two different approximations for the exchange - correlation functional .
the remainder of the paper is organized as follows .
details of the computational method as well as details of the experimental setup are outlined in section 2 .
the measured and calculated equations of state are presented in section 3 together with calculated ground state properties and elastic properties .
the electronic structure and the pressure variation of the band gap are discussed in section 4 .
finally , conclusions are given in section 5 .
the all - electron full - potential linear muffin tin orbital ( fp - lmto ) method@xcite is used to calculate the total energies and basic ground state properties of kmgf@xmath0 presented here . in this method ,
the crystal volume is split into two regions : non - overlapping muffin - tin spheres surrounding each atom and the interstitial region between the spheres .
we used a double @xmath3 spdf lmto basis ( each radial function within the spheres is matched to a hankel function in the interstitial region ) to describe the valence bands . in the calculations we included the 3s , 3p , 4s , 4p , and 3d bases for potassium , the 3s , 2p , 3p , and 3d bases for magnesium , and the 2s and 2p bases for fluorine .
the exchange correlation potential was calculated within the local density approximation ( lda)@xcite as well as the generalized gradient approximation ( gga ) scheme@xcite .
the charge density and potential inside the muffin - tin spheres were expanded in terms of spherical harmonics up to @xmath4=6 , while in the interstitial region , they were expanded in plane waves , with 14146 waves ( energy up to 156.30 ry ) included in the calculation .
total energies were calculated as a function of volume for a ( 16 16 16 ) k - mesh containing 165 k - points in the irreducible wedge of the brillouin zone and were fitted to the birch equation of state@xcite to obtain the ground state properties .
the elastic constants were obtained from the variation of the total energy under volume - conserving strains , as outlined in refs . and .
the high - pressure x - ray diffraction measurements used a sample of polycrystalline kmgf@xmath0 prepared by the solid state reaction method from high purity constituent materials as described elsewhere in several reports@xcite .
diffraction patterns collected at ambient temperature and pressure showed a cubic ( @xmath1 ) symmetry with a cell parameter a = 4.0060(2 ) for kmgf@xmath0 which closely agrees with earlier reports@xcite .
high pressures were generated by a merrill - bassett type diamond - anvil cell ( dac ) .
an 185-@xmath5 m sample chamber was formed in a rhenium metal gasket with a pre - indention of 60-@xmath5 m thickness .
the powder sample was loaded in the gasket with a few ruby grains and silicone fluid as pressure transmitting medium@xcite .
diffraction experiments were performed at the 16id - b undulator beam line of the high pressure collaborative team ( hpcat ) of the advanced photon source ( aps ) .
a monochromatic x - ray beam with a wavelength of 0.4218 was focused down to a size of 30 x 30 @xmath5 m .
diffraction images were collected with an image plate detector for an exposure time of 10 sec .
the distance between the sample and the detector and the inclination angle of the image plate were calibrated using a ceo@xmath6 standard .
the two dimensional images were subsequently integrated to one dimensional diffraction patterns using the fit2d software@xcite .
the cell parameters were obtained by analyzing the diffraction patterns with the jade software package and the p - v data obtained was fitted with a second - order birch - murnaghan@xcite equation of state .
the standard ruby fluorescence technique and the newly proposed ruby pressure scale of holzapfel@xcite were used to obtain the pressures in the sample chamber .
[ figxrd ] +
powder x - ray diffraction patterns collected at several pressures are shown in figure 1 . on compression ,
the diffraction patterns remain unchanged up to 40 gpa , except for the shifts of diffraction lines caused by the decreasing lattice constant .
this implies that no structural transformations occur up to 40 gpa in kmgf@xmath0 .
figure 2 shows the measured equation of state of kmgf@xmath0 and compares it with theoretical curves calculated within the lda and gga .
the better theoretical description is obtained with the lda which is somewhat surprising , since usually the gga provides an improvement over lda . at low pressures
the lda volume is slightly smaller than the experimental one , while the situation reverses at high pressures , _
i. e. , _ altogether the lda predicts kmgf@xmath0 to be stiffer than experimental observations .
gga on the other hand greatly overestimates the equilibrium volume at ambient pressure , which is the main reason for the poor agreement with experiment .
if the gga curve is scaled througout the pressure range with the error in equilibrium volume at @xmath7 , nearly perfect agreement is found with experiment ( not shown ) .
the lattice constant and bulk modulus measured in the present work as well as values calculated within the lda and gga approximations are given in table i. results from earlier experimental and theoretical works are quoted for comparison .
the bulk modulus obtained in our experiments @xmath8 = 71.2(2 ) gpa with @xmath9 = 4.7(3 ) compares well with other experimental results listed in table 1 and also with namgf@xmath0 reported recently by liu et al@xcite ( @xmath10 gpa ) .
the lattice constant obtained within the lda is 1.1 % lower than the experimental value , while the corresponding bulk modulus is 22% higher than the experimental value , which is the usual kind of accuracy of lda .
however the calculated lda lattice constant from the present work agrees quite well with the experimental work when compared to the earlier fp - lapw(lda ) calculations in which the reported lattice constant is 2.4% lower than the experimental value@xcite .
the lda bulk modulus obtained from the present calculation agrees quite well the published fp - lapw(lda ) results .
when comparing the results obtained within gga , the lattice constant is 1.9 % higher than the experimental value , whereas our results for the bulk modulus is within the spread of the experimental data .
this truly excellent agreement regarding the bulk modulus is , however , a bit fortuitous .
since the calculated equilibrium volume is overestimated with gga ( and underestimated with lda ) , an error solely depending on the error in volume is introduced in the calculated bulk modulus .
therefore , we recalculated the bulk modulus also at the experimental volume in a manner similar to our earlier work@xcite ( see table i ) .
we find that this diminishes the discrepancies between the lda and gga results , as expected .
in addition , the lda bulk modulus now becomes _ smaller _ than the gga one for kmgf@xmath0 , and both functionals are seen to actually overestimate the bulk modulus , by approximately 14 % ( lda ) and 34 % ( gga ) .
the present experiments on kmgf@xmath0 relate to recent experiments performed for namgf@xmath0 and alloys of namgf@xmath0 and kmgf@xmath0 .
the crystal chemistry of na@xmath11k@xmath12mgf@xmath0 and namgf@xmath0 was studied in detail at ambient and at high pressures by zhao et al .
namgf@xmath0 undergoes a reversible phase transition from orthorhombic ( @xmath13 ) to tetragonal ( @xmath14 ) and then to cubic structure ( @xmath1 ) upon compression .
these phase transitions require either compositional changes , by increasing the k concentration to 40% , or changing temperature or pressure .
the structural changes in these perovskites are due to octahedral tiltings and shortening of mg - f bonds compared to the cubic phase . a direct transformation from orthorhombic to cubic structure in namgf@xmath0 ,
however , requires a very high temperature ( 1038 k ) .
moreover , the transition temperature is reported to increase with pressure .
the temperature dependence of the crystal structure of kmgf@xmath0 was recently investigated by neutron powder diffraction by wood et al.@xcite from 4.2 k to 1223 k , and the cubic symmetry was found to be stable throughout this temperature range .
the thermal expansion as well as the atomic displacement parameters obtained in their experiments show that the f ions behave less anistropically than in namgf@xmath0 at such high temperatures . on comparing these results with the present high - pressure diffraction experiments on kmgf@xmath0
, one may speculate that application of pressure alone would not be sufficient to induce structural changes in kmgf@xmath0 as the cubic phase is very stable .
such a phase transformation if any , would require either application of very high temperature or a composition change in the system to achieve changes in the order parameters .
asbrink et al.@xcite have studied single crystals of the transition metal bearing perovskite kmnf@xmath0 , which is isostrucural to kmgf@xmath0 , under high pressure and observed a cubic - to - tetragonal phase transition at a critical pressure of 3.1 gpa . on combining these results , phase transitions from the cubic symmetry
may be expected with a combination of composition change , temperature and pressure in kmgf@xmath0 .
a systematic study on the octahedral tilting and order parameters with other dopant compositions and the effect of external thermodynamical variables are further required to understand the phase stability of kmgf@xmath0 .
[ eos ] in the pressure range from 0 - 40 gpa .
the experimental datapoints are marked by circles , while theoretical results obtained with lda and gga are shown as full ( blue ) and dashed ( red ) curves .
volumes are given relative to the experimental equilibrium volume @xmath15 .
, title="fig:",width=377 ] + the elastic constants of kmgf@xmath0 calculated within lda and gga are listed in table 2 where they are also compared to experimental results as well as earlier calculations .
the lda overestimates all of the @xmath16 , @xmath17 and @xmath18 elastic constants by between 10% and 22% compared to experiment.@xcite the elastic constants obtained within gga are much closer to the experimental values than are the lda results .
for instance , both @xmath16 and @xmath17 are within the experimental spread .
of course , the elastic constants also depend sensitively on the volume , and , therefore , the same argument as for the bulk modulus can be applied here .
we have , however , refrained from recalculating all the elastic constants with the volume correction , but wish to mention that the excellent agreement between experiment and the gga elastic constants should be interpreted with care .
another point of caution is the fact that the calculated values pertain to 0 kelvin , while experiments are performed at room temperature .
finite temperature generally tends to reduce the elastic constants because of thermal expansion . using the calculated elastic constants we calculated the anisotropy factor @xmath19 .
we find an @xmath20 for lda and @xmath21 for gga .
the experimental result is 1.05 , measured at room temperature@xcite , which is closer to but lower than the gga value .
however , the anisotropy factor is found to decrease as the temperature is lowered@xcite .
a simple relationship , which empirically links the plastic properties of materials with their elastic moduli was proposed by pugh@xcite .
the shear modulus @xmath22 represents the resistance to plastic deformation , while the bulk modulus @xmath23 represents the resistance to fracture .
a high @xmath24 ratio is associated with ductility whereas a low value corresponds to brittle nature .
the critical value which separates ductile and brittle materials is around @xmath25 , i.e. if @xmath26 the material behaves in a ductile manner , otherwise the material behaves in a brittle manner .
frantsevich@xcite , in a similar fashion has suggested @xmath27 as the critical value separating brittle and ductile behavior . in the case of kmgf@xmath0
the calculated value of @xmath24 is 1.5 within lda and 1.4 within gga , hence classifying this material as brittle .
pettifor@xcite suggested that the angular character of atomic bonding in metals and compounds , which also relates to the ductility , could be described by the cauchy pressure @xmath28 . for metallic bonding
the cauchy pressure is typically positive . on the other hand , for directional bonding with angular character ,
the cauchy pressure is negative , with larger negative pressure representing a more directional character .
these correlations have been verified for ductile materials such as ni and al that have typical metallic bonding , as well as for brittle semiconductors such as si with directional bonding@xcite . in the ionic compound kmgf@xmath0 ,
the calculated cauchy pressure is -10 gpa within lda and -15 gpa within gga , in good agreement with the nonmetallic characteristics of kmgf@xmath0 .
table 3 presents sound velocities as derived from the calculated elastic constants@xcite .
the calculated sound velocities agree quite well with the experiments , in particular for the gga values , which is a consequence of the somewhat fortuitous good agreement between the measured and gga calculated elastic constants .
the calculated electron band structure of kmgf@xmath0 is shown in figure 3 with the ensuing density of states in figure 4.the valence bands consist of the f p bands with a gap of 7.24 ev to the conduction band , which is dominated by k states .
the lda bands are almost identical , however with a gap of only 6.95 ev .
the gap increases almost linearly with compression , at the rate @xmath29 the conduction band minimum occurs at the @xmath30 point , while the valence band maximum occurs at the r - point @xmath31 .
the largest occupied energy level at the m - point @xmath32 is marginally lower ( by @xmath33 ev ) than the valence band maximum at r , and it remains lower throughout the pressure range studied here .
[ figband ] ( using gga , at the experimental lattice constant ) .
the zero of energy is set at the position of the conduction band minimum . , title="fig:",width=377 ] + [ figdos ] ( using gga , at the experimental lattice constant ) .
the zero of energy is set at the position of the conduction band minimum .
the partial projections onto the spheres of k , mg and f are shown with dotted , dashed and dash - dotted lines , respectively , while full line gives the total density of states .
units are electrons per ev and per formula unit . , title="fig:",width=377 ] +
in conclusion , we find that pure cubic kmgf@xmath0 is very stable under high compression . from our analysis
we also find that it is a brittle system and an indirect gap insulator whose gap increases with pressure g. v , v. k , a. d and b. j acknowledge vr and ssf for financial support and snic for providing computer time .
one of the authors ( r.s.k ) is greatly indebted to his departed colleague yongrong shen for help in collecting the data .
the authors gratefully acknowledge the use of the hpcat facility supported by doe - bes , doe - nnsa , nsf , and the w. m. keck foundation .
hpcat is a high - pressure collaborative access team among the carnegie institution , lawrence livermore national laboratory , the university of nevada las vegas , and the carnegie / doe alliance center .
we thank dr .
maddury somarazulu and other hpcat staff for technical assistance .
this research was supported from the u.s .
department of energy cooperative agreement no
. de - fc52 - 06na26274 with the university of nevada las vegas . .
calculated lattice constants ( in ) , bulk modulus @xmath8 ( in gpa ) and its pressure derivative @xmath9 , of kmgf@xmath0 at the theoretical equilibrium volume compared with the experiment and other theoretical calculations .
the bulk moduli have been calculated both at the experimental and theoretical volume ( @xmath34 and @xmath35 , respectively ) [ cols="^,^,^,^,^,^ " , ] | the high - pressure structural behaviour of the flouroperovskite kmgf@xmath0 is investigated by theory and experiment .
density functional calculations were performed within the local density approximation and the generalized gradient approximation for exchange and correlation effects , as implemented within the full - potential linear muffin - tin orbital method . _ in situ _ high - pressure powder x - ray diffraction experiments were performed up to a maximum pressure of 40 gpa using synchrotron radiation .
we find that the cubic @xmath1 crystal symmetry persists throughout the pressure range studied .
the calculated ground state properties the equilibrium lattice constant , bulk modulus and elastic constants are in good agreement with experimental results . by analyzing the ratio between the bulk and shear modulii ,
we conclude that kmgf@xmath0 is brittle in nature . under ambient conditions ,
kmgf@xmath0 is found to be an indirect gap insulator with the gap increasing under pressure . |
in recent years , high spatial resolution observations of clusters of galaxies by _ chandra _ revealed the existence of sharp discontinuities of x - ray intensity called cold fronts ( e.g. , * ? ? ? * ; * ? ? ? * ) . their existence is challenging because in high temperature plasma
, high thermal conductivity quickly smears such a discontinuity in temperature .
@xcite pointed out that heat conductivity across the front should be reduced from the classical spitzer value , @xmath0 @xcite by several orders of magnitude .
magnetic fields enable such suppression .
cold fronts may also be subject to the kelvin - helmholtz ( k - h ) instability .
@xcite suggested that turbulent magnetic fields stretched along the cold front will reduce the growth rate of the k - h instability at the interface between a subcluster plasma and an ambient plasma .
the magnetic field strength in the core of clusters of galaxies can be the order of @xmath1 ( e.g. , * ? ? ?
* ; * ? ? ?
@xcite summarized the observations of the field strength in several regions of abell 3667 by using different techniques .
they showed that its strength in the cluster core is about @xmath2 , and that magnetic fields are tangled on scales of roughly @xmath3 .
let us estimate the heat conduction time scale in cluster of galaxies .
it has been suggested that the heat conduction in cluster of galaxies can be very efficient ( e.g. , * ? ? ?
* ) because the coulomb mean free path , @xmath4 , is large . here
, @xmath5 is the number density . the time required for heat to diffuse by conduction across a length @xmath6 is given by @xmath7 , where @xmath8 is the density and @xmath9 is the spitzer conductivity .
since the thickness of the cold front is much smaller than @xmath3 , they will disappear in a time scale shorter than @xmath10 .
if magnetic fields exist , however , the characteristic length of the heat exchange across magnetic field lines is reduced to the larmor radius , @xmath11 .
thus , the heat conduction across the magnetic field is almost suppressed .
earlier numerical studies of cold fronts ( e.g. , * ? ? ?
* ) did not include magnetic fields and heat conduction .
@xcite first reported the results of two - dimensional ( 2d ) magnetohydrodynamic ( mhd ) simulations of cold fronts including both magnetic fields and anisotropic heat conduction .
they showed that magnetic fields are stretched and elongated ahead of the front .
since these magnetic fields suppress the heat conduction across them , the contact discontinuity ( a cold front ) between the cold subcluster plasma and the hot ambient plasma can be maintained .
the cold subcluster is wrapped by magnetic fields and protected from being heated by the heat conduction . in this paper , we focus on the three - dimensional ( 3d ) effects . in the 3d case ,
since magnetic fields compressed in front of the cold front can expand in the third direction or slip out the front surface , the strength of the magnetic fields in this region may be reduced compared to that in the 2d case .
we would like to show that even when we include these 3d effects , cold fronts can be maintained .
we simulated the time evolution of a cluster plasma in a frame comoving with the subcluster .
the basic equations are as follows : @xmath12 @xmath13 = -\mbox{\boldmath $ \nabla$}{\it p } + \frac{(\mbox{\boldmath $ \nabla$ } \times \mbox{\boldmath $ b$ } ) \times \mbox{\boldmath $ b$}}{4 \pi } - \rho \mbox{\boldmath $ \nabla$ } \psi,\ ] ] @xmath14 @xmath15 } \\ & & \lefteqn { + \mbox{\boldmath$\nabla$$\cdot$ } \left [ \left(\frac{1}{2 } \rho v^{2 } + \frac{\gamma p}{\gamma -1 } \right ) \mbox{\boldmath$v$ } + \frac{1}{4 \pi } ( -\mbox{\boldmath $ v$ } \times \mbox{\boldmath$b$ } ) \times \mbox{\boldmath $ b$ } - \kappa_{\parallel } \mbox{\boldmath $ \nabla_{\parallel}$ } t \right ] = - \rho \mbox{\boldmath $ v$ $ \cdot$ $ \nabla$ } \psi , } \nonumber % \end{equation}\end{aligned}\ ] ] where @xmath8 , @xmath16 , @xmath17 , @xmath18 , and @xmath19 are the density , velocity , pressure , magnetic fields , and gravitational potential , respectively .
we use the specific heat ratio @xmath20 .
the subscript @xmath21 denotes the components parallel to the magnetic field lines .
we assume that heat is conducted only along the field lines .
we solved ideal mhd equations in a cartesian coordinate system @xmath22 by a modified lax - wendroff method with artificial viscosity .
the heat conduction term in the energy equation is solved by the implicit red and black successive overrelaxation method ( see * ? ? ? * for detail ) . the radiative cooling term is not included .
the units of length , velocity , density , pressure , temperature , and time in our simulations are @xmath23 , @xmath24 , @xmath25 , @xmath26 , @xmath27 , and @xmath28 , respectively .
figure [ fig : f1 ] shows the initial density distribution . solid lines and arrows show magnetic field lines and velocity vectors in the @xmath29 plane , respectively .
we assume that a spherical isothermal low - temperature ( @xmath30 ) plasma is confined by the gravitational potential of a subcluster .
the low - temperature plasma is embedded in the less - dense @xmath31 uniform , hot @xmath32 plasma . here
the subscripts `` in '' and `` out '' denote the values inside and outside the subcluster , respectively .
we assume that the density distribution of the subcluster is given by the @xmath33-model profile @xcite , @xmath34^{-3\beta/2}\ ] ] where @xmath35 , @xmath36 , the maximum density @xmath37 , and the core radius is @xmath38 .
the subcluster is initially in hydrostatic equilibrium under the gravitational potential fixed throughout the simulation .
we assume that ambient plasma initially has a uniform speed with mach number @xmath39 , where @xmath40 is the ambient sound speed .
the mach number with respect to the sound velocity inside the subcluster is @xmath41 .
table [ tbl1 ] shows model parameters .
an important parameter is the plasma beta ( @xmath42 ) defined as the ratio of the ambient gas pressure to the magnetic pressure . when @xmath43 , the magnetic field strength is @xmath44
. models hc2 ( 2d ) and hc3 ( 3d ) are non - magnetic models with isotropic heat conduction , and model mc2a ( 2d ) and mc3a ( 3d ) are models with magnetic fields ( @xmath45 ) and anisotropic heat conduction . model mc2b
( 2d ) is a model with weak magnetic fields ( @xmath46 ) and anisotropic heat conduction . model mc3b
( 3d ) is a model with weaker magnetic fields ( @xmath47 ) .
we note that the initial magnetic fields are parallel to the @xmath48-axis and perpendicular to the motion of the subcluster .
when magnetic fields exist , heat is conducted only in the direction parallel to the field lines . along the magnetic fields
, the heat conductivity is taken to be that of the spitzer conductivity , @xmath49 .
meanwhile the conductivity across the field lines is @xmath50 .
the conduction time scale along the field lines is @xmath51 . for boundary conditions ,
the left boundary at @xmath52 is taken to be a fixed boundary , and other boundaries are taken to be free boundaries where waves can be transmitted .
.models and parameters .
heat conduction is isotropic in models hc2 and hc3 . in other models
, heat conducts only along magnetic fields .
[ cols="^,^,^,^,^,^,^",options="header " , ]
figure [ fig : f2 ] shows the magnetic fields for model mc3a at @xmath53 . solid curves and color contours show the magnetic field lines and density distribution , respectively .
the subcluster plasma is wrapped by the magnetic field lines .
ahead of the subcluster , the magnetic field lines are almost parallel to the contact surface between the subcluster plasma and the ambient plasma .
we can identify magnetic field lines slipping in the @xmath54-direction .
figure [ fig : f3 ] shows isocontours of magnetic field strength in the @xmath29 plane for models mc3a ( left , 3d ) and mc2a ( right , 2d ) at @xmath53 , respectively .
arrows show velocity vectors .
we note that the initial magnetic field lines are parallel to the @xmath48-axis .
the magnetic field strength is enhanced ahead of the subcluster via compression .
compared to the 2d model ( right ) , magnetic fields in the 3d model are weaker ahead of the subcluster but their strength is enhanced at the tail of the subcluster because the plasma flows advect the magnetic fields toward the tail and they pile up in the tail . the left panel of figure [ fig : f4 ] shows the distribution of magnetic field strength at @xmath53 along the @xmath55-axis .
the solid curve shows their strength for model mc3a ( 3d , @xmath45 ) .
the dotted line and the long - dashed line show their strength for model mc2a ( 2d , @xmath45 ) and mc2b ( 2d , @xmath56 ) , respectively . their strength in front of the cold front
is enhanced more than a factor 3 from the initial state .
the right panel in figure [ fig : f4 ] shows the time evolution of the total magnetic energy integrated in the whole simulation region .
the magnetic energy increases gradually due to magnetic field compression around the subcluster . .
color contours show the logarithmic density distribution .
the solid curves show magnetic field lines.,width=283 ] between 3d ( left ) and 2d ( right ) .
the left and right panels show the result for models mc3a and mc2a at @xmath29 plane , respectively .
arrows show the velocity vectors .
, width=510 ] .
the solid curve shows magnetic field strength for model mc3a ( 3d , @xmath45 ) along the @xmath55-axis . the dotted curve and the long - dashed curve
show the strength for model mc2a ( 2d , @xmath45 ) and mc2b ( 2d , @xmath56 ) along the @xmath55-axis , respectively . the dashed line and dash - dotted line show the initial magnetic field strength for models with @xmath45 and @xmath56 , respectively .
right : time evolution of the magnetic energy ( @xmath57 ) integrated in the whole simulation region .
the increase with respect to the initial kinetic energy , @xmath58 , is plotted .
the solid line , dotted line , and long - dashed line are for models mc3a
( 3d , @xmath45 ) , mc2a ( 2d , @xmath45 ) , and mc2b ( 2d , @xmath56 ) , respectively.,width=510 ] let us compare the results for model mc3a ( with heat conduction ) and model hc3 ( without heat conduction ) .
we show in figure [ fig : f5 ] distributions of the projected density , temperature , and x - ray intensity obtained by integrating these quantities in the @xmath54-direction .
x - ray intensity are visualized from simulation results as logarithm of the thermal bremsstrahlung emissivity , @xmath59 . in model
hc3 ( lower panels ) ,
heat is conducted isotropically .
since the conduction time scale across the cold front with width @xmath60 @xcite is very short ( @xmath61 ) , the cool component inside the subcluster is heated by the hot ambient plasma and evaporates rapidly . meanwhile in model mc3a ( top panels ) , heat is conducted only in the direction parallel to the field lines . as we showed in
[ sec : res1 ] , since the subcluster is wrapped by magnetic field lines , the cool component inside the subcluster is protected from heating .
we can see this protection remarkably in the temperature distributions ( middle panels ) .
thus the discontinuity remains sharp .
figure [ fig : f6 ] shows the temperature profiles at @xmath62 along the @xmath55-axis .
the left panel shows those for models mc3a ( solid line ) and hc3 ( dotted line ) .
the right panel shows those for models mc2a ( solid line ) , mc2b ( dash - dotted line ) , and hc2 ( dotted line ) . in both panels ,
the dashed line shows the initial temperature profile .
it is obvious that magnetic fields along the front enables the cold front to exist for @xmath63 .
we can also identify shock waves .
numerical results reproduced the bow shock produced by supersonic motion upstream of the cold front observed in a3667 ( see figure [ fig : f5 ] ) .
the interaction of the subcluster plasma with the ambient plasma generates another shock propageting inside the subcluster .
this shock exists at @xmath64 in models mc3a and mc2a ( solid lines ) in figure [ fig : f6 ] , and also visible in the left panel of figure [ fig : f4 ] .
these shocks drive the adiabatic expansion of the subcluster and cools the subcluster plasma .
such behaviors are consistent with those in 2d models @xcite . along the @xmath54-axis .
the top panels show results for model mc3a ( with magnetic fields and heat conduction ) .
the bottom panels show results for model hc3 ( without magnetic fields and with heat conduction).,width=510 ] along the @xmath55-axis .
the left panel shows the results for models mc3a ( solid line ) , mc3b ( dash - dotted line ) , and hc3 ( dotted line ) . the right panel shows the results for models mc2a ( solid line ) , mc2b ( dash - dotted line ) , and hc2 ( dotted line ) . the dashed lines in both panels show the initial temperature distribution.,width=510 ]
we have shown that magnetic fields enable the existence of cold fronts by suppressing the heat conduction across the front .
previous hydrodynamic simulations ( e.g. , * ? ? ?
* ) could reproduce the cold fronts because they neglected the heat conduction .
when heat conduction is included , cold fronts disappear in time scales of @xmath65 unless the heat conduction is suppressed as we showed in models hc2 and hc3 ( see figures [ fig : f5 ] and [ fig : f6 ] ) .
magnetic fields much smaller than that in model mc3a can suppress heat conduction .
we confirmed that weak magnetic fields for model mc3b ( @xmath66 ) can suppress the heat conduction across the front as long as they cover the front .
even when large - scale coherent magnetic fields do not exist , turbulent magnetic fields can suppress the heat conduction across the front ; they are stretched and elongated parallel to the front ( see also * ? ? ?
* ; * ? ? ?
@xcite suggested that magnetic fields are important for the existence of cold fronts because they can suppress the growth of the k - h instability .
@xcite showed that in models without heat conduction and without magnetic fields , the growth of the k - h instability is not prominent and does not disrupt the cold front ahead of the moving subcluster . in the sides of the subcluster , the k - h instability creates eddies which detach and propagate downstream .
magnetic fields can suppress the formation of such eddies .
let us now discuss the possibility of the disruption of the subcluster plasma by magnetic fields .
@xcite presented the results of 3d simulations of moderate supersonic cloud motion in a magnetized interstellar medium .
they assumed that the cloud has uniform density .
they showed that magnetic fields accumulating ahead of the cloud triggers the rayleigh - taylor ( r - t ) instability and disrupts the cloud .
our numerical results , however , does not show the growth of such r - t instability .
these different behaviors are due to a difference of the density distribution of the cloud .
since the subcluster plasma in our simulations has a dense core because it is confined by the gravity , magnetic fields ahead of the clump can not deform the density distribution inside the subcluster .
we have shown that in the 3d case , since magnetic fields ahead of the subcluster slip along the contact surface between the subcluster plasma and the ambient plasma , the amplification of their strength is smaller than that in the 2d case
. this weaker magnetic field , however , is sufficient to suppress the heat conduction across the front . since the area of the contact surface is larger in the 3d case than that in the 2d case .
thus cold subcluster plasma can be heated more efficiently than in the 2d case .
we confirmed this by comparing results for model hc2 and hc3 ( figure [ fig : f6 ] ) . compared with model
( 2d , non - magnetic fields ) , the temperature profile for model hc3 ( 3d , non - magnetic fields ) shows that the cold subcluster is heated up earlier .
when magnetic fields exist , however , the difference becomes less prominent .
we thank t. yokoyama for developments of the coordinated astronomical numerical software ( cans ) which include 2d and 3d mhd codes including heat conduction .
the development of cans was supported by act - jst of japan science and technology corporation . this work is supported by the priority research project in graduate school of science and technology , chiba university ( p.i . , r. matsumoto ) .
numerical computations were carried out on vpp5000 at the astronomical data analysis center , adac , of the national astronomical observatory of japan .
m. johnston - hollitt , the magnetic field in a3667 , in : t. h. reiprich , j. c. kempner , n. soker , ( eds . ) , the riddle of cooling flows in galaxies and clusters of galaxies , url : http://www.astro.virginia.edu/coolflow/ , ( 2004 ) . | recent high resolution observations by the _ chandra _ x - ray satellite revealed various substructures in hot x - ray emitting plasmas in cluster of galaxies . for example , _ chandra _ revealed the existence of sharp discontinuities in the surface brightness at the leading edge of subclusters in merging clusters ( e.g. , abell 3667 ) , where the temperature drops sharply across the fronts .
these sharp edges are called cold fronts .
we present results of three - dimensional ( 3d ) magnetohydrodynamic simulations of the interaction between a dense subcluster plasma and ambient magnetized intracluster medium .
anisotropic heat conduction along magnetic field lines is included . at the initial state ,
magnetic fields are assumed to be uniform and transverse to the motion of the dense subcluster .
since magnetic fields ahead of the subcluster slip toward the third direction in the 3d case , the strength of magnetic fields in this region can be reduced compared to that in the 2d case . nevertheless , a cold front can be maintained because the magnetic field lines wrapping around the forehead of the subcluster suppress the heat conduction across them . on the other hand ,
when the magnetic field is absent , a cold front can not be maintained because isotropic heat conduction from the hot ambient plasma rapidly heats the cold subcluster plasma .
, , magnetohydrodynamics , plasmas , conduction , x - ray , magnetic fields , galaxy clusters 95.30.qd , 95.30.tg , 95.85.nv , 95.85.sz , 98.65.cw |
satellites of the giant planets in the solar system are typically categorized into regular and irregular moons .
regular moons orbit their host planet in circular , low inclination short period orbits ; irregular moons have highly inclined ( and even retrograde ) , typically eccentric and have longer period orbits .
similar to planet formation , the formation of satellite systems around giant planets in the solar system ( and in exoplanetary systems ) is difficult to study theoretically .
regular satellites are thought to have formed through collisional growth of smaller planetesimals in the circum - planetary disk , very similar to planet formation in the coagulation model for the terrestrial planets .
irregular satellites are typically thought to have formed elsewhere in the solar system ( as asteroid / kuiper - belt object like planetesimals ) , and only later be captured to become moons .
the various models suggested have major difficuties in explaining the origin of the irregular moons ( ( see * ? ? ?
* for a review ) ) .
more succesful recent capture models have been suggested ; these involve planet - planet interactions , and the existence of an additional third ice - giant planet in the solar system that have been later ejected from it @xcite .
an oversimplified but useful approach used to study satellite formation is to divide it into two stages , based on the importance of the effects of gas and the circum - planetary disk on the growing satellites .
the first stage lasts a few myr , until the dissipation of the gaseous circum - planetary disk , following the dissipation of the gaseous protoplanetary disk which fuels it . at the end of this stage large moons
should have formed @xcite in addition to and ( possibly many ) smaller moonlets .
the second stage begins following the gas dissipation . at this stage
the evolution is driven primarily by gravitational interactions and collisions between the moons . in comparison with the extensive study of the first stage of satellite - system formation in a gaseous disk , the dynamical relaxation through moon - moon scattering at the later stage
have scarcely been explored . here
we study this dynamical evolution using analytic techniques and n - body simulations of moon - moon scattering around giant planets , and suggest that it can play a major role in the build - up and sculpting of the architecture of satellite systems , both in the solar system as well as in ( not yet detected ) exomoon systems .
this paper is structured as follows .
we first provide a brief overview of the various processes seen in our n - body moon - moon scattering simulations , which are later discussed in more detail .
we introduce the n - body simulations used to study the satellite systems , and discuss the initial conditions following the dissipation of the gaseous circum - planetary disk , followed by a brief discussion of the stability of multi - moon systems .
we then provide an analytic context to dynamical relaxation of multi - moon systems .
we describe several examples of multi - moon systems studied through n - body simulations and present their evolution and outcomes , demonstrating the possible outcomes .
finally we discuss the evolution of multi - moon system through scattering and their implications for satellites systems as well as the populations of minor bodies in the solar system ( asteroids , trans - neptunian objects , comets etc ) .
dynamical evolution of a satellite system goes through various stages .
we will first review these stages , and then discuss them in more detail both analytically and through examples from n - body simulations . *
instability * : following the formation of the moons in an extended circum - planetary disk , they mutually perturb each other gravitationally . as long as a gaseous disk or a planetesimal disk exist , such disks might dissipate the perturbation and stabilize the system . after the disk is gone
, the mutual perturbations may destabilize the systems on short timescales , if their masses are large and/or the separations between them are small enough ( similar to the well studied planet - planet scattering case;e.g .
the perturbations can then grow and lead to orbit crossing between the various moons . here
we study only the evolution of satellite system at the onset of dynamical instability after the gaseous disk has been gone . * scattering , collsional growth and production of prograde irregular satellites : * in an unstable system the moons will scatter each other gravitationally , leading to orbital changes , exciting the eccentricity and inclinations of the satellites .
when the relative velocities between the moons become larger than the escape velocity from their surface , physical collisions between moons will occur , which might alleviate further growth of the moons and/or their possible disruption .
moons excited to high eccentricities may collide with the host planet .
alternatively , a moon can be scattered to a large distance extending beyond the stable region around the planet ( typically @xmath0 or even @xmath1 for prograde and retrograde orbits , respectively ) , and then be ejected from the system to become a runaway moon . through the scattering process moons can migrate in the disk and exchange positions to produce a mixed order , i.e. erasing the signature of the original moon ordering .
* secular evolution and the kozai - lidov `` barrier '' * : moons excited to high inclinations with respect to the orbital plane of the host planet around the sun , can become susceptible to large amplitude secular orbital evolution due to the tidal perturbations by the sun or by other moons ( so called kozai - lidov oscillations ) .
such secular evolution can drive the satellite orbits into extremely eccentric orbits , and eventually lead to their collision with another moon or the planet , or to their ejection from the system .
it is therefore difficult for moon scattering processes to directly lead to excitation of retrograde orbits from initially prograde orbits .
a similar limitation on eccentricities could also be seen in planet - planet scattering simulations of exo - planetary systems ( e.g. * ? ? ?
scattering to stable high inclination orbits is therefore limited by a kozai - lidov `` barrier '' .
* temporary and permanent re - capture of runaway moons and the origin of retrograde irregular moons * : following their ejection , runaway moons , now on helio - centric orbits , are likely to interact again with the host planet .
some can interact with the other solar system planets and/or protoplanetary disk and may obtain a stable helio - centric orbit .
others might strongly interact with the planet and even be ejected from the solar system .
runaway moons can also be re - captured temporarily into orbits around the planet even into retrograde orbits with respect to the orbit of the planet around the sun . such re - captured moons can potentially dissipate some of their kinetic energy through scattering and/or collisions with other moons orbiting the planet and/or tidal interaction with the planet ( i.e. somewhat similar to planet capture from wide eccentric orbits into close retrograde orbits in planet scattering simulations ; @xcite ) , and thereby be re - captured into permanent orbits .
we conjecture that such dissipative processes could produce retrograde irregular moons ( similar to the models envisioning capture of objects formed independently in the protoplanetary disk ) . here
we only model purely dynamical interactions and do not include dissipative processes ( beside direct collisions ) , which will be explored elsewhere .
throughout this paper we discuss the results of several n - body simulations ( using the mercury code @xcite ) of satellite systems . in following papers
we will describe the statistical results of a large sample of simulations through the study of a wide range of initial conditions .
the current paper focuses on presenting the basic processes and outcomes of moon - moon scattering process , and only presents results from a small number of simulations to provide explicit examples .
the results of these specific simulations are corroborated by the much larger set of simulations ( see payne & perets ; paper i ) , for detailed discussion of the extended n - body simulations ) .
table 1 provides the the initial conditions for these simulations .
we have used a single neptune like planet ( neptune mass and separation from the sun ) , orbited by a large number of moons ( 10 - 50 massive moons , and in some cases up to 1000 massless particles corresponding to low mass moons ) . in all simulation
the outermost moon is placed at a distance @xmath2 ( i.e. at stable orbits comparable to the outermost prograde satellites in the solar system ; @xmath3 is the planet hill radius ) and subsequent moons are placed internal to this moon at separations of @xmath4 mutual hill radii from each other . in all simulations shown here the moons
are put on co - planar orbits coinciding with the orbital plane of the planet orbit around the sun ; with only small random inclinations , distributed randomly between @xmath5 and 0 - 0.01 eccentricities .
though we show results of moon - moon scattering around a neptune like planet , results of similar satellite systems around other solar system like planets show similar results ( perets et al .
, in prep . ) .
[ cols="^,^,^,^,^,^,^,^",options="header " , ]
the first stage of moon formation has been explored by several groups who provided a plausible explanation for the origin of the regular moons in the a small ( typically @[email protected]@xmath8 ) circum - planetary disk .
studies of the formation of circum - planetary disks , however , suggest that such disks should extend up to a large fraction of the planetary hill radius ( up to approximately @xmath9 ; ( e.g. * ? ? ?
* ; * ? ? ?
* ) ) , and are not truncated at small separations , as was hitherto assumed ( e.g. * ? ? ?
* ; * ? ? ?
* ) ; based on older non - resolved simulations of the gaseous disks ; e.g. @xcite ) .
formation of regular moons can therefore potentially extend much beyond the region of the currently observed regular moons . given these results ,
the likely initial conditions for a satellite system is the formation of a multi - moon system extending up to a large fraction of @xmath8 .
consequently , the initial condition assumed throughout this study are the initial existence of a disk like population of moons extending up to 0.3 @xmath8 ( the typical stability region for gas planet moons ) .
a detailed study of the _ initial stages _ of regular moons formation in an extended circum - planetary disk is beyond the scope of this paper and will be discussed elsewhere . here
we focus on the _ dynamical evolution _ of moons in extended disks .
a general criteria for the stability of a multi - object keplerian system has not been found yet . for a system with two low mass planets on circular , nearly co - planar orbits ,
a stability criteria was found by @xcite and @xcite , who showed that such systems do not have a close encounter ( they are hill stable ) if their semi - major axes are separated by @xmath10\text{.}\label{eq : stability}\end{aligned}\ ] ] we can now scale this criteria for a satellite system around a planet , and consider two moons ( where the solar mass is replaced by the planetary mass , and @xmath11 , @xmath12 now refer to the satellite masses ) . for
a system of two moons with @xmath13to be stable we then require from eq .
[ eq : stability ] that @xmath14a_{1}.\label{eq : stability-2}\ ] ] from this relation one can find that some of the known pairs of neighboring regular moons in the solar system could have been dynamically unstable , if they were located further away than their current orbit ( e.g. in the irregular moons region ) , or if they were more massive then their current measured masses . in higher multiplicity systems
direct n - body simulations show that initially stable systems require a few mutual hill radii separation between the planets in order not to destabilize on a short timescale @xcite ; however no general stability criteria is known for @xmath15 objects orbiting a central massive object . in all the simulations of moon scattering we initialized all moons to reside at 4 - 6 mutual hill radii from their closest neighbor .
let us first consider a single encounter between two moons , occurring at some distance @xmath16 from the planet , assuming a fast encounter at the impulse approximation limit .
the kick velocity to a given moon ( assumed to be smaller , for simplicity ) through scattering by a moon of mass @xmath17 ( and radius @xmath18 ) is of the order of @xmath19 where @xmath20 is the impact parameter , and @xmath21 is the relative velocity between the moons , typically comparable to the velocity dispersion in the circum - planetary satellite disk . for a heated disk ,
the velocity dispersion is of the order of @xmath22 where @xmath23 is the height ratio of the disk , and @xmath24 is the keplerian velocity at distance @xmath16 from the planet ( cite ) , i.e. @xmath25 . comparing the kick velocity gained through a close approach ( comparable to a few times the satellite radius , @xmath26 i.e. @xmath27 , for which we can assume tidal effect are negligible ) to the keplerian velocity ( for a circular orbit ) around the planet at that position , @xmath28 we get the radius at which a mutual kick could significantly change the satellite orbit ( where we assume a constant disk height ratio , @xmath29 ) @xmath30 replacing the keplerian velocity with the escape velocity , we can similarly get the critical radius from which moons can be ejected from the system ( though in practice , due to the tidal perturbation by the sun moons ejected beyond @xmath31 would already become unbound from the system ) . for scattering by the most massive observed moons for each of the giant planets , we find that @xmath32 @xmath8 , taking the satellites ganymede , titan , titania and triton ( for jupiter , saturn , uranus and neptune , respectively ) .
it is therefore clear that even single strong encounters with such moons can significantly alter the orbits of companion moons and even eject them , if found beyond @xmath33 , which , for reasonable @xmath34 ratios includes most of the stable region around these planets . as mentioned above , no analytic criteria for instability of multi - planets
is currently known , but simulations of multi - planet systems show that such systems evolve into strong scatterings and orbit crossing when the mutual distance between planets are a few mutual hill radii or less . in our simulations
we typically separate each neighboring moon by 4 mutual hill radii , but we find that even larger separations ( 5 - 6 mutual hill radii ) eventually lead to strong scatterings ( see also paper i ) . the dynamical evolution of a disk of planetesimals has been extensively studied both numerically and analytically . making use of similar tools we would expect viscous stirring to excite the eccentricities and inclinations of the moons in circum - planetary disk . for a disk composed of @xmath35 equal mass moons , with mass @xmath36 , the evolution of the velocity dispersion in the disk , @xmath37 , with time would then go like ( following @xcite ; see also @xcite ) @xmath38 for a disk ( or ring ) centered at radius @xmath39 , with a radial width @xmath40 ; @xmath41 is the coulomb logarithm ( @xmath42 ) and @xmath43 is the keplerian orbital period at @xmath39 ; @xmath44 is a constant pre - factor ( @xcite find @xmath45 in their disk simulations ) .
effectively , as the disk evolves moons may be ejected or collide and therefore @xmath35 and/or @xmath36 should evolve with time . nevertheless , this should serve as a good approximation at the earlier stages of the evolution , before a significant number of ejections / collisions of moons occur .
the change in disk thickness would also affect the maximal effective distance between encounters , thereby affecting @xmath46 ; but the latter effect is logarithmically weak .
we therefore expect the evolution of a circum - planetary disk to be similar to that of a protoplanetary disk , at least at the early stages and far from the instability region where moons can be easily ejected , i.e. @xmath47 . in fig .
[ fig : viscous - stirring ] we show the evolution of a single component disk in our n - body simulation ( model 2 in table [ tab : simulation - models ] ) , and compare it with the simplified analytic results ( no fit ; predictions for two extreme @xmath44 values , @xmath48 are shown ) ; we assume a disk size of @xmath49 au around a neptune like ( mass and distance from the sun ) planet .
we show the evolution of the rms eccentricity and inclination ( @xmath50 ; @xmath51 ) . at later times ( @xmath52yrs ) the simple assumptions used in the model break down ( due to ejections / collisions which change the number density and the mass of the moons ) and the rms eccentricity of the moons evolve slightly slower than the analytic predictions ( not shown ) , dynamical evolution of a disk of moons around a neptune like planet .
we simulated a circum - planetary disk composed of 50 equal mass moons ( @xmath53 g ) initially distributed between 0.01 au and 0.2 au , with 4 mutual hill radii separation between each two neighboring moons ( model 2 ) .
shown is the evolution of the rms eccentricity ( upper solid line ) and inclination ( lower solid line ) compared with the analytic predictions ( dash - dotted and dashed lines correspond to @xmath54 , @xmath55 respectively ; see text , eq .
[ eq : sig - evol ] ) .
simulation results are smoothed with a window of 20k yrs for clarity .
the multi - moon system shows similar evolution to that of a protoplanetary planetesimal disk .
note that at later evolutionary times ( not shown here ) , the disk evolution differs since the tidal perturbation from sun become more significant for moons close to @xmath8 and/or at high inclinations , where significant secular evolution may occur . ] as discussed above , moon - moon scattering dynamically heats the satellite system , leading to moon orbit crossing and potential physical collisions and the coagulation of smaller moons into big moons . though collisions are included in our simulations ( assuming sticking spheres model for the collision ) , we focus on the late stages of satellite systems after the main stage of the moon formation , where only a small number of massive moons reside in the disk .
a more detailed study of satellite formation beyond the region of regular moons is beyond the scope of this study .
we note , however , that collisions do play a role in the satellite system evolution , and a large fraction of the satellite mutually collide .
[ fig : mass_growth ] shows an example of the mass function evolution of a satellite system ( initially with equal mass moons ) .
collisional mass growth evolution of satellites .
the mass of each satellite is shown as a function of time ; all satellites were initalized with the same mass ( model 2 ) , @xmath56 @xmath57(lines for different moons are moved up consecutively by @xmath58 for better view ) . of the 50 satellites at the begining of the simulation
, 24 survived to the end of the simulation at 60 myrs .
of these satellites the figure shows only the six which have grown in mass through collisions .
as can be seen one of these moons has been produced through the cummulative collision of 19 satellites . ] as illustrated in fig .
[ eq : sig - evol ] ( based on results from model 2 ) , moon - scattering excites the eccentricities ( and inclinations ) of the satellites , as well as changes their semi - major axis .
if the separation of a moon from its host planet ( at apocenter ) extends beyond the stability region around the planet , the tidal perturbations by the sun start to dominate its dynamical evolution , leading to fast large amplitude fluctuations in its orbital elements .
typically such evolution would , eventually , lead to the ejection of the moon from the system as its orbit extends beyond the hill radius . in some cases , however , scattering by other moons close to pericenter can decrease the apocenter distance of the moon from the planet and bring it back into a stable orbit .
we find that satellites can migrate back and forth throughout the stability region , but , as can be seen clearly in figs .
[ fig : orbital_evolution ] and [ fig : orbital_evolution-2 ] , they can not survive for long beyond @xmath59 .
the evolution of a satellite system is somewhat similar to that of an evaporating globular cluster in the tidal field of the galaxy .
relaxation processes lead to the contraction of the system into a tighter configuration accompanied by the ejection of moons either diffusively by slowly migrating beyond the stability region of the host planet , or through more rare strong encounters , which can scatter a moon from a stable inner orbit directly into a wide separation and unstable orbit . our simulations clearly demonstrate both these evaporation channels ( see fig .
[ fig : orbital_evolution ] ) .
generally we find that , similar to multi - planet scattering results ( e.g. * ? ? ?
* ) , at the end of our simulations typically only 1 - 3 massive moons survive , irrespective of the initial number of massive moons in the simulations ; this is corroborated by a large cohort of 3 - 10 massive moons simulations ( using initial moon masses comparable to those of the most massive solar system regular moons ) .
the rest of the moons typically either collide with the planet or are ejected from the system into heliocentric orbits , thereby forming a population of `` runaway '' moons ( see fig .
[ fig : orbital_evolution-2 ] for the evolution of such runaway moons ) .
the results of these large phase - space studies are discussed in a companion paper ( payne et al .
, in prep . )
. dynamical history of three different test - mass moons in model 3 , two of which are eventually ejected from the system due to interactions with the massive satellites .
bottom : evolution of apo - center .
the ( red ) dots show the slow evolution of the satellite orbit into the unstable zone .
the ( green ) pluses show a satellite ejection through a strong scattering from an inner stable orbit .
note that both ejected satellites were later temporarily captured .
the first ( red dots ) is recaptured and even evolves into stable small separation orbit at very high and even retrograde inclinations ( see bottom panel ) , before colliding with the planet . the ( blue ) rectangles show the evolution of a never ejected satellite .
middle : evolution of the satellite eccentricity .
note the high eccentricities at the post - repcature stage of the satellite ; eventually leading to the collsion of the first ( red dots ) satellite with the planet .
top : evolution of satellite inclinations .
note the retrograde orbit of the first recaptured ( red dots ) satellite . ]
two examples of n - body simulation results for the evolution of satellite system through moon - moon scattering .
the evolution captures self - excitation , scattering and ejections of moons in a system consisting of three triton - mass moon orbiting a neptune - like planet ( model 4 ) .
the simulations were initialized with three moons spaced at 3-mutual hill radii apart , with the outer - most moon positioned at 14 hill planetary hill radii .
the separation between the moons and the planet ( in units of the planetary hill radius ) is shown as a function of time ; also plotted is a horizontal line at 0.5 hill radii , at which to guide the reader to the approximate region where stable satellites can not survive . the eccentricities and inclinations of the surviving moons in such systems are excited to high eccentricities ( @xmath60 ) and inclinations ( @xmath61 radians ) . ]
two examples of n - body simulation results for the evolution of satellite system through moon - moon scattering .
the evolution captures self - excitation , scattering and ejections of moons in a system consisting of three triton - mass moon orbiting a neptune - like planet ( model 4 ) .
the simulations were initialized with three moons spaced at 3-mutual hill radii apart , with the outer - most moon positioned at 14 hill planetary hill radii .
the separation between the moons and the planet ( in units of the planetary hill radius ) is shown as a function of time ; also plotted is a horizontal line at 0.5 hill radii , at which to guide the reader to the approximate region where stable satellites can not survive . the eccentricities and inclinations of the surviving moons in such systems are excited to high eccentricities ( @xmath60 ) and inclinations ( @xmath61 radians ) . ]
it is well established that irregular satellites can be strongly affected by secular kozai - lidov ( kl ) evolution due to perturbations by the sun ( see * ? ? ? * and references therein ) .
such evolution leads to periodical / quasi - periodical changes in the inclinations and eccentricities of the satellites .
the amplitude of such oscillations can become large when the inclinations with respect to the orbit of the host planet around the sun are larger than @xmath62(somewhat smaller for non - zero eccentricity orbits ) .
in particular , excitation of large eccentricities can lead to mutual orbit crossing of the various moons and eventually to strong scattering between them .
the extremely high eccentricities potentially reached through this process also make the moons susceptible to physical collisions with the planet and/or to be affected by strong tidal interactions ( at pericenter ) . at apocenter
such moons might be ejected from the system if they approach the instability region at apocenter ( approximately 0.3 - 0.4 of the planetary hill radius ) . for these reasons ,
moons scattered into high ( mutual ) inclination ( with respect to the planet orbit around the sun ) are not likely to survive long before they collide with the planet / other moon or are ejected from the system .
such evolution depletes moons at high inclinations ; the lifetime of moons at such inclinations is short ; they can not achieve high inclinations close to @xmath63 , which would allow them , through additional scattering ( or secular evolution ) obtain retrograde orbits .
rather , they are either scattered into at most @xmath64 and are then scattered back to lower inclinations or they are destroyed ( collision or ejection ) .
we note , however , that kl evolution is very sensitive to other perturbations affecting the pericenter precession of the orbits , and can be easily quenched due to competing interactions ( e.g. the combined perturbations of the other moons alter the pure kl evolution due to the sun ; tidal interaction and/or the oblateness of the planet can also alter such evolution , at least for moons achieving small pericenters ; see also @xcite for a similar discussion in the context of planet - planet scattering ) .
though long term stability of high inclination orbit is difficult , we nevertheless find many moons do chive retrograde orbits throughout their orbital evolution , sometimes even for long period of time ( thousands of years ) .
moon - moon scattering can therefore produce not only prograde irregular satellites , but also retrograde irregular ones .
analysis of their dynamical history reveals that all moons showing this behavior have been , at earlier times , ejected from the system , and have later been temporarily recaptured into retrograde orbits ( fig .
[ fig : orbital_evolution ] show a typical example of such evolution ) .
such a temporary capture , if followed by a dissipative process , could produce irregular moons ; indeed this is the basic mechanism suggested to produce irregular moons in the moon - capture scenario .
our results therefore suggest that in - situ formation of moons followed by scattering can lead to similar outcomes as the capture scenarios , without alluding to capture of externally formed planetesimals , i.e. irregular moons might be recaptured moons , in addition to , or instead of being captured asteroids / kbos . in our simulations
we do not account for any dissipation from e.g. interaction with gas or tidal friction during close encounters with the host planet .
scattering by other moons can also stabilize the orbit of such recaptured runaway moons to become permanently captured .
nevertheless , we find no satellites that survived in retrograde orbits more that few thousands years .
in fact , we find that all the moons which eventually collided with the host planet in our simulations did so only after they have been ejected and had been temporarily recaptured . in fig .
[ fig : captured_incs ] we show the inclination distribution of such satellites prior to their collision .
this suggests that introducing tidal interaction with the planet can have a major part in potentially stabilizing the orbits of these moons , and potentially produce the retrograde satellites we see today .
indeed , it had been shown that the tidal effects are essential for producing retrograde planets in planet - scattering simulations ( e.g. payne et al . ,
submitted ) .
the potentially promising effects of adding such a dissipative process to the moon - scattering simulations is beyond the scope of this paper and will be discussed elsewhere .
distribution of inclinations for potentially tidally captured / dissipated moons ( in model 3 ; that @xmath65 au ) , at the point they where they would have ( strongly ) tidally interacted with the planet ( pericenter distance smaller than twice the planetary radius ) .
the addition of tidal interactions could potentially dissipate sufficient energy as to recapture such moons , and quench any further excitation of the eccentricity which would have otherwise lead to their collision with the planet . ]
the discovery of hundreds of exo - planetary systems in the last two decades have provided new perspectives on the possible architecture of planetary systems .
in contrast to the solar system with its co - planar configuration of planets in nearly circular orbits , many exo - planetary systems present planets on highly inclined ( with respect to the spin of their host star ) and/or on highly eccentric orbits .
planets are thought to form in a protoplanetary disk which could naturally account for the existence of planets on co - planar circular orbits .
the new discoveries of eccentric / inclined planetary orbits gave rise to models in which following the formation of multiple planets , they mutually perturbed and scatter each other through gravitational interactions .
such dynamical excitation eventually leads to the ejection of most of the planets , and the survival of typically 1 - 3 planets on eccentric and inclined orbits , with orbital properties consistent with observations .
regular moons of the solar system gas giants have been suggested to form in a circum - planetary disk around their host planet , in an analogue process to the formation of the solar system planets and exoplanets observed in co - planar , circular orbits .
as demonstrated here , massive moons ( comparable in mass to the regular moons ) which formed on co - planar circular orbits beyond the region of the regular moons can scatter each other and/or scatter a population of low mass satellites ( massless particles in our simulations ) into eccentric and inclined configurations thereby producing irregular - like moons , serving as the satellite analogue for eccentric / inclined exo - planetary systems .
our main results can be summarized as follows 1 .
most of the massive moons are ejected , typically leaving behind 1 - 4 surviving moons on stable configurations , irrespective of the initial number of moons , very similar to the results of planet - planet scattering experiments in multi - planet scattering studies ( e.g. @xcite ) .
the rest of the moons are either ejected , collide with one each other ( potentially leading to further satellite growth ) or collide with the planet .
2 . the surviving massive moons can obtain high eccentricities and inclinations , similar to those of prograde irregular moons observed in the solar system .
however , none of the massive moons obtain a stable retrograde orbit .
a non - negligible fraction of the small moons ( massless particles ) in our simulations ( @xmath66% ) are ejected from the system but are then temporarily recaptured , for times between a few years to a few thousand of years . of these recaptured moons
@xmath67 % have retrograde inclinations just before they collide with the planet .
we therefore conclude that multiple moon systems can behave very similarly to multiple planet systems , and can produce satellite systems with high eccentricity and high inclination _ prograde _ irregular satellites .
we conjecture that moons observed to temporarily attain retrograde orbits during their evolution can potentially be stabilized through dissipative processes ( such as strong interaction with a disk / other moons or the planet ) and may also produce stable _ retrograde _ irregular moons , somewhat similar to the observed behavior of planet - scattering simulations which consider tidal interactions ( e.g. payne et al . ,
submitted ) .
note that currently no massive irregular moons beside triton are observed in the solar system ; though this may serve as an important constraint on the moon - moon scattering scenario ever happening in the solar system , it is possible that massive irregular moons were later removed through dissipative processes not studied here ( e.g. migration in the circum - planetary disk into close separations to the host planet , tidal interaction with the planet , or strong disrupting collisions ) . the introduction of dissipative processes in moon - moon scattering simulations will be studied in detail elsewhere .
finally , we note that moon - moon scattering mechanism can produce large planetesimals in heliocentric orbits that were originally formed as moons , and were later ejected from their host planet .
such planetesimals might still be observed today as asteroids / kuiper belt objects , and may have unique kinematic and/or peculiar properties in term of composition and/or structure compared to the background heliocentric formed planetesimal population , due to their different origin . | gas giant planets in the solar system host large satellite systems with multiple regular and irregular moons .
regular moons revolve around their host planet in circular , low inclination short period orbits , and are thought to form in - situ through coagulation processes .
in contrast , irregular moons have highly inclined ( and even retrograde ) , typically eccentric and long period orbits around their host planet .
irregular moons are therefore often thought to have formed as unbound objects in helio - centric orbits that were later captured to their current orbits around the planet
. however , such capture scenarios require fine tuned conditions and/or encounter difficulties in producing irregular moon populations around all gas giants . here
we study the possibility that regular moons form in - situ outside the currently observed regular - moon regime , and dynamically evolve through mutual moon - moon scattering ( as well as by secular evolution due to perturbations by the sun ) .
we find that such evolution can excite the satellites into high eccentricities and inclinations .
we find that moons are either ejected from the host planet to become runaway moons , or stay bound and become prograde orbiting irregular moons with inclined and eccentric orbits around their host planet .
ejected moons , unbound to the planet , can later be temporarily re - captured by the host planet even at retrograde orbits .
such moons are eventually re - ejected from the system or collide with the planet , at least in the absence of dissipative processes ( e.g. collisions with existing bound moons , a debris disk or through tidal interactions with the host planet ) , not currently modeled .
uncaptured runaway moons may eventually be ejected from the solar system , or be captured into stable helio - centric orbits and contribute to the populations of asteroidal or trans - neptunian objects .
such scenarios are relevant both for the gas - giant satellites in the solar system and for the dynamical evolution of exomoons . |
the standard model ( sm ) of particle physics provides an excellent fit to precision electroweak experiments , provided the mass @xmath0 of the as - yet unobserved higgs boson is rather light @xcite , @xmath1 however , the sm suffers from the hierarchy problem : the electroweak scale is sensitive to higher energy scales occurring in nature , via a quadratically - divergent contribution to the higgs mass parameter coming dominantly from virtual top quarks .
the divergence must be cut off by new physics at some energy scale @xmath2 , which , in the absence of unnatural fine - tuning of parameters , should also be rather low .
this has been a major motivation for the ongoing construction of the large hadron collider ( lhc ) .
a quantitative measure of the fine - tuning constraint may be given as @xcite @xmath3 where @xmath4 and the amount of fine - tuning is roughly one part in @xmath5 ; in a natural theory , @xmath5 should be of order unity and the new physics will thus be accessible at the lhc .
the minimal supersymmetric extension of the standard model ( mssm ) , in which the quadratic divergence is elegantly cut off by the superpartners of the top quarks ( the stops ) , is already unnatural in this sense : the stops must have masses of many hundreds of gev in order to push the mass of the lightest higgs boson above the empirical lower bound obtained from direct searches at lep @xcite and the fine - tuning is at least a few _ per cent_.
however , the electroweak ( [ ewsm ] ) and fine - tuning ( [ ftsm ] ) constraints are changed in a theory with an extended higgs sector . indeed , in such a model , the higgs mass eigenstates that couple significantly to gauge bosons ( and are thus constrained by electroweak precision tests ) need not coincide with the eigenstates which are sensitive to @xmath2 .
the overlap between the different eigenstates ( and _ ergo _ the constraints ) is necessarily complete only in the theory with a single higgs boson , _ viz . _
what is more , the individual constraints generalizing ( [ ewsm ] ) and ( [ ftsm ] ) are themselves changed in an extended higgs sector and may also relax the constraint on the scale of new physics .
recently , barbieri and hall @xcite showed that even the simplest extension of the sm , the ( non - supersymmetric ) two higgs doublet model , exhibits regions of parameter space with ` improved naturalness ' , in which the combined electroweak and fine - tuning constraints are relaxed , allowing the scale of new physics to be as large as 2 tev or more .
this leads to the rather pessimistic conclusion that the new physics , whatever it may be , could lie beyond the reach of the lhc @xcite . in the following ,
we exhibit a supersymmetric model of this type @xcite , the so - called ` fat higgs model ' @xcite .
the existence of such a model implies that , even if no direct evidence for supersymmetry is seen at the lhc , it can not be ruled out as the solution to the hierarchy problem .
models with supersymmetry necessarily have multiple higgs doublets and are , as such , candidates for having improved naturalness . in supersymmetric models , the new physics ( which cuts off the quadratic divergence in the higgs mass parameters )
is provided by superpartners .
the dominant divergence comes from top quarks , for which the contribution to the up - type higgs mass parameter @xmath6 is given by @xmath7 where @xmath8 is the top quark yukawa coupling , @xmath9 is the stop mass , and @xmath10 is the cut - off . in a theory which remains perturbative up to ,
say , the gut scale ( such as the mssm or the nmssm ) , this relation implies a two - fold suppression of the natural stop mass .
firstly , on the left - hand side , @xmath6 can not be far above the weak scale , because it is related to the electroweak _ vev _ by a factor given roughly by the higgs quartic couplings , which grow in the ultra - violet and so can not be much larger than unity at the weak scale if they are to remain perturbative all the way up to the gut scale .
secondly , on the right - hand side , the logarithm ( in which @xmath10 is the gut scale ) is large enough to cancel the loop factor .
these two effects conspire in such a way that the superpartner scale is essentially the same as the weak scale , up to a factor of order unity .
the obvious way out is to discard the requirement that the theory remain perturbative up to the gut scale . then @xmath6 ( and the higgs masses )
can be rather large , and the logarithm ( in which @xmath10 is now the strong - coupling scale ) is small .
what is more , successful unification of gauge couplings is not contingent upon quartic couplings remaining perturbative @xcite : it is possible to allow the theory to go through a calculable supersymmetric strong - coupling transition , whilst retaining gauge - coupling unification at higher energies ( just as , for example , electroweak unification at around a hundred gev is not spoiled by the couplings of the low energy effective chiral lagrangian for qcd becoming strongly - coupled at a gev or so ) . below the strong coupling scale , the fat higgs model @xcite contains the higgs doublet chiral superfields @xmath11 and @xmath12 of the mssm together with a singlet superfield @xmath13 .
the superpotential contains the terms @xmath14 we expect @xmath15 at the strong coupling scale @xmath10 on the basis of naive dimensional analysis @xcite , while at a lower energy scale @xmath16 , the coupling decreases according to the renormalization group .
following @xcite , we take the quartic coupling @xmath17 to be large compared to the gauge couplings @xmath18 and @xmath19 and neglect the @xmath20-terms in what follows .
we have already given a qualitative explanation of how the fine - tuning constraint is alleviated in a strongly - coupled model .
a rigorous discussion is given in @xcite .
the fit to the electroweak data is described in @xcite .
constraints on the @xmath21 and @xmath22 parameters from electroweak data .
the elliptical contours , centered on the best - fit values ( + ) , enclose the 68 , 90 , 95 and 99 % confidence regions .
the grey area shows the model contribution for points in parameter space with @xmath23 tev and satisfying the @xmath24 and @xmath25 constraints . for comparison ,
the thick line indicates the sm contribution with higgs masses between 100 gev ( at the origin ) and 1 tev in increments of 100 gev . from @xcite.,width=283 ] fig .
[ st ] shows various confidence regions for @xmath21 and @xmath22 ( relative to a sm higgs reference mass of 100 gev ) , together with the model contributions in regions of parameter space with natural stop mass above 2 tev with the fine - tuning parameters @xmath26 , for the sake of argument ) , and satisfying the @xmath24 and @xmath25 constraints . in fig .
[ hb ] , we plot the higgs masses _ versus _
@xmath27 for points with @xmath23 tev which are also compatible with electroweak constraints at the 95 @xmath28 c.l .
as is clear from fig .
[ st ] , large values of @xmath29 are compatible with electroweak precision tests because there are large , positive contributions to @xmath22 , coming from the other higgs scalars .
can an extended higgs sector of this type be discovered at the lhc ?
both neutral higgs scalars lie above the threshold for the ` gold - plated ' decay @xmath30 but the width is shared with the decay to @xmath31 , which has a large background @xcite .
thus , it is not clear that either @xmath32 or @xmath33 will necessarily be discovered in this scenario .
k. inoue , a. kakuto , h. komatsu and s. takeshita , prog .
* 68 * , 927 ( 1982 ) ; k. inoue , a. kakuto , h. komatsu , and s. takeshita , prog .
. phys . * 67 * , 1889 ( 1982 ) ; r. a. flores and m. sher , ann . phys . * 148 * , 95 ( 1983 ) ; h. e. haber and r. hempfling , phys .
lett . * 66 * , 1815 ( 1991 ) ; y. okada , m. yamaguchi and t. yanagida , prog .
. phys . * 85 * , 1 ( 1991 ) ; j. r. ellis , g. ridolfi and f. zwirner , phys .
* b257 * , 83 ( 1991 ) . | we present a supersymmetric model of electroweak symmetry - breaking exhibiting improved naturalness , wherein the stop mass can be pushed beyond the reach of the large hadron collider without unnatural fine tuning .
this implies that supersymmetry may still solve the hierarchy problem , even if it eludes detection at the lhc .
address = rudolf peierls centre for theoretical physics , university of oxford , 1 keble rd .
, oxford ox1 3np , uk , altaddress = merton college , oxford ox1 4jd , uk address = rudolf peierls centre for theoretical physics , university of oxford , 1 keble rd .
, oxford ox1 3np , uk |
the link between the irreversible phenomena governed by macroscopic equations such as the navier - stokes equations and the microscopic reversible dynamics of the atoms and molecules is a fundamental problem . in this context
, it has been shown that typical many - body systems of interacting particles present a chaotic dynamics @xcite .
this microscopic chaos develops a sensitivity to initial conditions over a time scale of the order of the intercollisional time of the atoms and molecules .
this sensitivity to initial conditions is characterized by positive lyapunov exponents which are the rates of exponential separation between some reference and perturbed trajectories of the system .
the sensitivity to initial conditions results into a huge dynamical randomness characterized by a positive kolmogorov - sinai ( ks ) entropy per unit time given by the sum of positive lyapunov exponents if the system is at equilibrium : @xmath0 an identity known as pesin s equality @xcite .
this dynamical chaos provides an efficient mechanism of randomization of the different observable quantities such as the microscopic currents associated with the transport properties . for a hamiltonian - type microscopic dynamics
, a connection can be established between dynamical chaos and the transport properties thanks to the escape - rate formalism @xcite . in this formalism , the gap between the kinetic time scale of chaotic properties and the hydrodynamic time scale of transport properties is bridged by linking the transport coefficients to _ differences _ between chaotic properties . in this formalism , the following formula has been derived for viscosity @xcite @xmath1 where the difference between the sum of positive lyapunov exponents and the ks entropy is nonvanishing because the system is here under _ nonequilibrium _ conditions .
these nonequilibrium conditions select the trajectories of the many - body system which do not escape out of a phase - space region specific to the transport property of interest .
this phase - space region is defined by requiring that the helfand moment associated with the transport property remains bounded in an interval of extension @xmath2 . in the limit @xmath3 ,
the nonequilibrium condition is progressively relaxed and the sum of positive lyapunov exponents as well as the ks entropy approach their equilibrium value satisfying pesin s equality ( [ pesin ] ) . under nonequilibrium conditions , pesin s equality is not satisfied and the difference gives the rate of escape of trajectories out of the aforementioned phase - space region @xcite .
this region contains a fractal repeller @xmath4 composed of trajectories which escape neither in future nor in past .
the escape rate is characteristic of this fractal repeller and is related to the transport coefficient , leading to the formula ( [ esc.eq ] ) .
the escape - rate formalism has already been applied to the transport property of diffusion @xcite as well as to reaction - diffusion processes @xcite .
the purpose of the present paper is to apply the escape - rate formalism to viscosity .
the system we use as a vehicle of our study is a minimal model of viscosity previously analyzed by bunimovich and spohn @xcite .
the minimal models of transport are of special interest because they are the simplest possible models already presenting a positive and finite transport coefficient .
it is known that a minimal model should contain only one particle for diffusion , two particles for viscosity , and three particles for heat conduction @xcite . for viscosity , we therefore consider here the model composed of two hard disks moving on a torus and undergoing elastic collisions . in a previous paper @xcite
, we have described this model and some of its properties for a hexagonal and a square geometries .
our aim is here to compute viscosity thanks to the escape - rate formalism and to show the equivalence with the results of the green - kubo and einstein - helfand formulas already obtained in the previous paper @xcite . in the present paper ,
we compute the viscosity by using the chaotic and fractal properties of the repeller .
we use a variant of eq .
( [ esc.eq ] ) in which the difference between positive lyapunov exponents and ks entropy is given in terms of the lyapunov exponents and the partial fractal dimensions of the fractal repeller .
indeed , the fractality of the repeller is a corollary of its chaoticity so that its fractal dimensions are related to its ks entropy .
accordingly , the knowledge of the partial dimensions allow us to evaluate the ks entropy .
the paper is organized as follows . in sec .
[ escape ] , we develop the escape - rate formalism for shear viscosity . in sec .
[ repeller ] , we present the fractal repeller of viscosity in the two - disk model , which we compare with the fractal repeller of diffusion in the lorentz gas . in this way , we show that a specific fractal repeller is associated with each transport property .
finally , the chaotic and fractal properties are described in sec .
[ chaos ] where we compute the viscosity coefficient from the positive lyapunov exponent and the fractal dimension of the repeller .
conclusions are drawn in sec .
[ conclusions ] .
in the previous paper @xcite , we have shown that the shear viscosity coefficient can be computed with the einstein - helfand formula @xcite @xmath5 ^ 2 \rangle \ ; , \label{einstein}\ ] ] where the helfand moment is defined by @xmath6 \label{helfand}\ ] ] for a system of @xmath7 particles of position @xmath8 and momentum @xmath9 moving in a domain delimited by periodic boundary conditions . as explained in the previous paper @xcite
, the particles must satisfy the minimum image convention , which requires the presence of the extra terms in the helfand moment ( [ helfand ] ) involving the jumps @xmath10 of the particles to fulfil the periodic boundary conditions .
@xmath11 are the times of the jumps .
@xmath12 is the momentum at the time @xmath11 of the jump . in eq .
( [ einstein ] ) , the average @xmath13 is taken in the equilibrium microcanonical ensemble for which @xmath14 the central object of the escape - rate formalism is the fractal repeller which is composed of the phase - space trajectories for which the helfand moment fluctuates forever ( in the future and the past ) within some interval : @xmath15 these trajectories are exceptional because the helfand moment escape out of this interval for almost all the trajectories .
therefore , the repeller has a vanishing probability measure in the phase space albeit it is typically composed of a nonenumerable set of trajectories .
the repeller thus typically forms a fractal in the phase space @xcite .
we set up a first passage problem of the helfand moment by introducing absorbing boundaries at @xmath16 .
these absorbing boundaries in the space of variations of the helfand moment correspond to equivalent absorbing boundaries in the phase space of the system . in the phase space ,
the absorbing boundaries delimit a domain which contains the fractal repeller .
we consider a statistical ensemble of initial conditions taken inside this domain and we run their trajectories .
when a trajectory reaches the absorbing boundaries it escape out of the domain and it thus removed out of the statistical ensemble . under the forward time evolution ,
the remaining trajectories belong to the stable manifolds of the repeller . under the backward time evolution ,
the remaining trajectories belong to the unstable manifolds of the repeller . under both the forward and backward time evolution ,
the remaining trajectories belong to the repeller itself which is the intersection of its stable and unstable manifolds @xcite . for a typical chaotic dynamics , almost all trajectories escape out of the domain after some time so that the repeller as well as its stable or unstable manifolds are fractal objects .
these fractals can be generated by allowing the escape of trajectories over a long but finite time interval . over a finite time
, there remain a sizable set of trajectories , which progressively reduces to the fractal as the time interval becomes longer and longer .
the number of trajectories in the set ( or statistical ensemble ) decays with time . typically , the decay is exponential and characterized by the so - called escape rate @xmath17 .
the escape rate @xmath17 can be evaluated by solving the first passage problem of the helfand moment by introducing absorbing boundaries at @xmath18 . indeed , the einstein - helfand equation ( [ einstein ] ) shows that the helfand moment performs a diffusive - like random walk .
accordingly , the helfand moment can be considered as a random variable @xmath19 for which the probability density @xmath20 obeys to a diffusion - type equation @xcite : @xmath21 where the role of the diffusion coefficient is played by the shear viscosity ( [ einstein ] ) itself . at the absorbing boundaries ,
the probability density must satisfy the absorbing boundary conditions : @xmath22 the solution of the diffusion - type equation ( [ fokker - planck ] ) with the boundary conditions ( [ abc.p ] ) is given by @xmath23 with @xmath24 and where the coefficient @xmath25 depend on the initial probability density .
the number @xmath26 of trajectories remaining between the absorbing boundaries at the current time @xmath27 is related to the probability density by : @xmath28 after a long time , the escape is dominated by the smallest decay rate , @xmath29 , which can therefore be identified with the escape rate @xmath17 . in this way , we obtain the _ escape rate _ as a function of @xmath2 : @xmath30 this result is obtained by using the diffusion - type equation ( [ fokker - planck ] ) which is expected to hold over spatial distances larger than the mean free path of the particles .
therefore , the parameter @xmath2 must be sufficiently large so that the helfand moment is in a diffusion regime and eq .
( [ fokker - planck ] ) holds .
the shear viscosity coefficient can thus be obtained from the escape rate which depends on the parameter @xmath2 of separation between the absorbing boundaries as @xmath31 in the following we call eq .
( [ visc - esc ] ) the escape - transport formula . at the microscopic level of description ,
the escape rate is controlled by the fractal repeller @xmath4 which is composed of all the trajectories which satisfy the condition ( [ abc ] ) under forward and backward time evolution .
the repeller is the support of a natural invariant probability measure .
this invariant measure is natural because it is generated by the dynamics and can be approximated by a statistics based on the trajectories remaining within the absorbing boundaries after a long but finite time .
the dynamics on the fractal repeller is characterized by positive lyapunov exponents and a ks entropy , both evaluated with respect to the natural invariant measure of the repeller .
if the dynamics is unstable some lyapunov exponents are positive .
if the dynamics is chaotic the ks entropy is positive . on a repeller ,
the sum of positive lyapunov exponents differs from the ks entropy and the difference gives the escape rate @xmath32 of the repeller @xmath4 : @xcite @xmath33 if we combine this result from dynamical systems theory with eq .
( [ visc - esc ] ) , we obtain the chaos - transport formula ( [ esc.eq ] ) for viscosity as originally derived by dorfman and gaspard @xcite .
an equivalent formula can be obtained which involves the partial fractal dimensions of the repeller instead of the ks entropy .
indeed , the fractal character of the repeller is a direct consequence of the escape of trajectories so that the ks entropy is no longer equal to the sum of lyapunov exponents but to @xmath34 where the coefficients are the partial information dimensions of the repeller associated with each unstable direction of corresponding lyapunov exponent @xmath35 @xcite .
these partial dimensions satisfy @xmath36 so that the ks entropy is in general smaller than the sum of positive lyapunov exponents . accordingly
, the escape rate can be expressed as @xmath37 in terms of the partial codimensions defined as @xmath38 combining with eq .
( [ visc - esc ] ) , the shear viscosity is given by @xmath39 in the limit @xmath3 , the lyapunov exponents reach their equilibrium values @xmath40 , while the codimensions vanish typically as @xmath41 if transport is normal .
if we introduce the quantities @xmath42 eq .
( [ visc - codim ] ) provides a decomposition of the viscosity coefficient on the spectrum of lyapunov exponents such as @xmath43 typically , the escape is most important in the most unstable direction corresponding to the maximum lyapunov exponent @xmath44 .
therefore , the repeller is more depleted in the most unstable direction and the corresponding partial dimension @xmath45 is lower than the further ones .
this reasoning suggests that a typical behavior is @xmath46 for @xmath3 if the maximum lyapunov exponent @xmath44 is well defined .
this is precisely the case in two - degree - of - freedom systems such as the two - disk model where the chaos - transport formula reduces to @xmath47 where @xmath48 is the unique positive mean lyapunov exponent and @xmath49 the corresponding codimension which should be understood as the partial information codimension of the unstable manifolds of the fractal repeller given in terms of the partial information dimension by young s formula @xcite @xmath50 it is known that the partial information dimension of the repeller is well approximated by the partial hausdorff dimension if the escape rate is small enough and if ruelle s topological pressure does not present a discontinuity .
this last condition is fulfiled if the system does not undergo a dynamical phase transition , which is the case in the finite - horizon regimes of sinai s billiard which controls the dynamics of both the lorentz gas and the two - disk model @xcite . under these conditions
, we can replace the partial information codimension by the partial hausdorff codimension in the chaos - transport formula and obtain the viscosity as @xmath51 in the limit @xmath3 , the lyapunov exponent converges to its equilibrium value so that we can also write : @xmath52 under similar conditions as eq .
( [ visc - hcodim ] ) .
in this section , our purpose is to display the fractal repeller associated with viscosity in the two - disk model and to compare it with the fractal repeller of diffusion in the lorentz gas in order to show that they are different and therefore specific to each transport property .
the two - disk model has been studied by bunimovich and spohn who proved thanks to a central limit theorem that viscosity is positive and finite in this minimal model @xcite . in the previous paper @xcite ,
we have considered the two - disk model in the hexagonal geometry which we shall use in the following .
we showed in ref .
@xcite that the dynamics reduces to a sinai billiard in the center - of - mass frame and that the helfand moment ( [ helfand ] ) with @xmath53 is then given by @xmath54 \ ; , \label{helfand2}\ ] ] where @xmath55 are the coordinates of the relative position of both disks and @xmath56 the canonically conjugated relative momentum .
the jumps happen when the trajectory of sinai s billiard crosses the hexagonal boundary .
if the trajectory crosses the side of label @xmath57 the trajectory is reinjected at the opposite side so that the jump in position is given by the lattice vector @xmath58 corresponding to the side @xmath57 .
a fractal repeller is defined by considering all the trajectories such that their helfand moment satisfies the conditions @xmath59 where the parameter @xmath2 should be large enough .
the stable manifolds of the fractal repeller can be visualized by plotting the initial conditions of trajectories satisfying the conditions ( [ abc - visc ] ) over a long time interval extending forward in time .
these initial conditions are taken on the disk of sinai s billiard .
the initial conditions are specified by the angle @xmath60 of the initial position and the angle @xmath61 that the initial velocity makes with a vector which is normal to the disk at the initial position ( see fig . [ variables - particules ] ) .
the initial conditions are plotted in the birkhoff coordinates ( @xmath60,@xmath62 ) .
figure [ fractal - visc - tot ] depicts such the fractal composed of the stable manifolds of the repeller for viscosity in the two - disk model .
we provide evidence that the set is fractal by zooming successively on it in figs .
[ fractal - visc - zoom1 ] and [ fractal - visc - zoom2 ] , where the self - similarity of the repeller clearly appears .
let us take a section across the repeller in fig .
[ fractal - visc - tot ] at @xmath63 . taking the escape time of the corresponding trajectory
, we have obtained the escape - time function depicted in fig .
[ temps - esc - visc-8 g ] .
the time for the trajectory to escape out of the phase - space region corresponding to the interval ( [ abc - visc ] ) is infinite if the trajectory belongs to the stable manifold of a trajectory of the repeller .
indeed , this trajectory is then asymptotic to a trajectory which does not escape .
accordingly , the escape - time function has vertical asymptotes on the stable manifolds of the repeller .
since the repeller is fractal the vertical asymptotes are not enumerable , which explains the behavior in fig .
[ temps - esc - visc-8 g ] .
diffusion of a tracer particle in the hard - disk periodic lorentz gas has been studied with the escape - rate formalism in ref .
@xcite . in this lorentz gas
, the tracer particle undergoes elastic collisions on hard disks forming a triangular lattice . in a unit cell of the lattice
, the dynamics also reduces to sinai s billiard .
the energy of the tracer particle is conserved as well as the phase - space volumes .
sinai and bunimovich have proved that the dynamics is ergodic and mixing and that the diffusion coefficient is positive and finite in the finite - horizon regime @xcite . for diffusion , the associated helfand moment is simply given by one of the coordinates @xmath55 of position of the tracer particle @xcite .
an escape process is associated with diffusion by setting up a problem of first passage of the tracer particle at some absorbing boundaries .
if we consider the @xmath64-coordinates , the tracer particle does not escape as long as the following condition is satisfied : @xmath65 the absorbing boundary conditions are therefore defined at @xmath66 . with these absorbing boundaries ,
the system is called an _ open lorentz gas _ @xcite .
the trajectories trapped within the interval ( [ abc - diff ] ) form a fractal repeller as shown in ref . @xcite .
in order to compare with the fractal repeller of viscosity , we can plot the fractal repeller of diffusion in a similar way as here above for viscosity . here
again , we plot all the initial conditions of trajectories remaining within the interval ( [ abc - diff ] ) over a long forward time interval .
these initial conditions are plotted in the same birkhoff coordinates ( @xmath60,@xmath62 ) of a disk around the coordinate @xmath67 in the lorentz gas .
the set of the selected initial conditions approximate the stable manifolds of the fractal repeller .
we zoom successively on this fractal in figs .
[ fractal - diff - zoom1 ] and [ fractal - diff - zoom2 ] , which provides evidence of its self - similarity . as a consequence , the repeller is also fractal .
the fractal dimension of the repeller is related to the diffusion coefficient of the lorentz gas and its lyapunov exponent , as shown in refs .
@xcite .
we observe that the two fractal repellers associated respectively with diffusion ( see fig .
[ fractal - diff - tot ] ) and viscosity ( see fig . [ fractal - visc - tot ] )
are different .
indeed , although the global structure is similar , the trajectories belonging to the different repellers are not the same .
to convince us of this difference , we take some examples of trajectories . in fig .
[ repeller - comparaison1 ] , we have a periodic trajectory bouncing between two disks in the billiard .
this trajectory belongs to the repeller associated with diffusion since the position @xmath64 is bounded and satisfied ( [ abc - diff ] ) .
however , the viscosity helfand moment of this trajectory does not satisfy the condition ( [ abc - visc ] ) so that it does not belong to the repeller of shear viscosity . with eq .
( [ helfand2 ] ) , we see that , in one direction , both @xmath68 and @xmath69 are positive . therefore , the contribution at this passage is positive for the helfand moment . in the other direction , both @xmath68 and @xmath69 are negative but the product @xmath70 is also positive .
accordingly , the helfand moment increases forever on this trajectory which , therefore , does not belong to the repeller associated with shear viscosity . on the other hand
, we can observe the opposite case .
figure [ exemple - diff - non - repeller ] depicts an example of trajectory escaping from the interval ( [ abc - diff ] ) although its helfand moment of viscosity remains between the absorbing boundary conditions ( [ abc - visc ] ) .
the repellers associated with different transport properties are therefore different . in this subsection , we show that the shear viscosity can be obtained from the escape rate of the repeller by using the escape - transport formula ( [ visc - esc ] ) .
we consider a sequence of repellers with larger and larger values of the parameter @xmath2 .
the escape rate @xmath32 is numerically evaluated for each repeller by computing the decay of the number @xmath26 of trajectories still within the interval ( [ abc - visc ] ) at current time and by extracting the escape rate from the exponential decay .
the escape rate is observed to behave as @xmath71 and the shear viscosity coefficient is then obtained with eq .
( [ visc - esc ] ) .
figures [ superposition - visc - helf - taux - hex ] and [ superposition - visc - helf - taux - carre ] depict the viscosity directly computed from the escape rate and compared with the values obtained by the einstein - helfand formula in ref .
@xcite , respectively in the hexagonal and square geometries . as in the previous paper @xcite , we consider reduced viscosities defined by @xmath72 we observe in figs .
[ superposition - visc - helf - taux - hex ] and [ superposition - visc - helf - taux - carre ] the excellent agreement between both methods .
in this section , we compute the shear viscosity coefficient in terms of the chaotic and fractal properties of the repeller by using the chaos - transport formula ( [ visc - hcodim ] ) which related the viscosity to the lyapunov exponent and the hausdorff codimension of the repeller of viscosity . in sinai s billiard which controls the reduced dynamics of the two - disk model , the elastic collisions between the disks are defocusing .
this induces a dynamical instability of the trajectories which is characterized by the lyapunov exponents .
these exponents are the rates of exponential separations between a reference orbit and infinitesimally close orbits .
since the dynamics of sinai s billiard is symplectic and volume - preserving in the four - dimensional phase space , the lyapunov exponents spectrum is ( @xmath73 ) so that their sum is vanishing .
one of the lyapunov exponents vanishes because of the absence of exponential separation in the direction of the flow .
another one corresponding to the direction perpendicular to the energy shell equals zero because of energy conservation .
there exists a method to calculate the positive lyapunov exponent by considering the motion of a front of particles accompanying the reference particle and issued from the same initial position but with different initial velocities @xcite . because the dynamics is defocusing ,
this front is expanding .
locally on the reference orbit @xmath74 the front ( called the unstable horocycle ) is characterized by a curvature @xmath75 or , equivalently , by its radius of curvature @xmath76 . thanks to
this method explained in detail in ref .
@xcite , we have computed the positive lyapunov exponent as a function of the density of the system ( in the hexagonal and square geometries ) .
the equilibrium values of the lyapunov exponent are obtained by running a trajectory in sinai s billiard without absorbing boundaries and by averaging over a long time interval .
the resulting numerical values are depicted in figs .
[ lyap - stretching - hex ] and [ lyap - stretching - carre ] . in the chaos - transport formula ( [ visc - hcodim ] )
, the lyapunov exponent has to be evaluated for the trajectories belonging to the fractal repeller .
the statistical average is here carried out for the natural invariant probability measure concentrated on the fractal repeller .
this invariant measure defines a nonequilibrium state for the motion . as aforementioned ,
the natural invariant measure is generated by the dynamics itself .
accordingly , the lyapunov exponent is numerically computed by averaging over a statistical ensemble of trajectories which has not yet escaped after a long but finite time
. this ensemble can be as large as wished by increasing the number of initial conditions . in this way
, we can calculate the nonequilibrium values of the lyapunov exponent . in table
i , we present a comparison between the equilibrium lyapunov exponent @xmath77 without absorbing boundary conditions ( as depicted in fig .
[ lyap - stretching - hex ] and [ lyap - stretching - carre ] and the nonequilibrium lyapunov exponent @xmath78 evaluated over a nonequilibrium measure which has the fractal repeller as support .
the difference between these exponents is small and of the order of the escape rate , in agreement with the results of ref .
@xcite for the disordered lorentz gas .
0.8 cm ' '' '' height1pt ' '' '' 0.3 cm table i. values of the characteristic quantities of chaos for different densities @xmath79 in the hexagonal system : @xmath77 is the equilibrium lyapunov exponent for the closed system . the following quantities characterize the fractal repeller for viscosity with @xmath80 : @xmath81 is the nonequilibrium lyapunov exponent of the repeller , @xmath82 its ks entropy , @xmath17 its escape rate , @xmath83 its partial information codimension , and @xmath84 its partial hausdorff codimension . 0.5 cm ' '' '' 0.2 cm @xmath85 0.05 cm ' '' '' ' '' '' height1pt 1 cm in order to determine the viscosity by the chaos - transport formula ( [ visc - hcodim ] ) , we need to determine the partial hausdorff codimension @xmath84 of the fractal repeller . the corresponding dimension @xmath86 is the hausdorff dimension of the vertical asymptotes of the escape - time function depicted in fig .
[ temps - esc - visc-8 g ] .
its values range in the interval @xmath87 .
the hausdorff codimension can be obtained using the following numerical algorithm developed by the group of maryland @xcite .
we consider an ensemble of pairs of trajectories starting from initial conditions @xmath88 differing in a value @xmath89 . the time taken by the trajectories to escape out of the system
is given by the escape - time function in fig .
[ temps - esc - visc-8 g ] .
the pair is said to be _
uncertain _ if the trajectories and their helfand moment present at least one of the following conditions : ( i ) both trajectories follow paths that differ by the successive passages through the cell boundaries , that is , if we associate to each trajectory a symbolic sequence ( @xmath90 ) which gives the labels of the cell boundaries across which the successive passages occur , and both sequences are different ; ( ii ) one of both trajectories has its helfand moment which reaches the fixed absorbing boundaries ( [ abc - visc ] ) when the helfand moment of the other one still remains within these limits .
if the pair does not present one of these conditions it is called _
certain_. the fraction @xmath91 of uncertain pairs in the initial ensemble is known to behave as the power @xmath92 giving the hausdorff codimension as its exponent .
derivations of this result can be found elsewhere @xcite .
this method has been already used in refs .
@xcite .
we have here applied the maryland algorithm to obtain the hausdorff codimension of the fractal repeller of viscosity .
table i compares the partial hausdorff codimension with the partial information dimension in particular cases .
we observe that both codimensions take very close values as expected . by varying the parameter @xmath2 ,
we have obtained the shear viscosity thanks to the chaos - transport formula ( [ visc - hcodim ] ) .
these values are plotted in figs .
[ visc - helf - taux - dim - hex_dens ] and [ visc - helf - taux - dim - carre_dens ] for the hexagonal and square geometries , respectively .
we consider the shear viscosity coefficient @xmath93 in the hexagonal geometry and the element @xmath94 of the viscosity tensor in the square geometry .
the values obtained with the chaos - transport formula ( [ visc - hcodim ] ) are compared with the values obtained by the escape - transport formula ( [ visc - esc ] ) and those by the einstein - helfand formula ( [ helfand ] ) obtained in the previous paper @xcite .
the agreement between the three formulas is excellent , which confirms the theoretical results .
in the present paper , we have applied the escape - rate formalism to the computation of shear viscosity in the two - disk model by bunimovich and spohn @xcite .
the escape - rate formalism implies the appearance of a fractal repeller associated with viscosity .
we have numerically generated the fractal repeller associated with viscosity in this model and we have provided evidence for its fractal character . using the chaos - transport formula of the escape - rate formalism , we have been able to evaluate the shear viscosity from the positive maximum lyapunov exponent and the hausdorff codimension of the fractal repeller of viscosity .
the values obtained by using the chaos - transport formula for shear viscosity have been compared with the values obtained by other methods based on the einstein - helfand formula , which is equivalent to the green - kubo formula as shown in our previous paper @xcite .
an excellent agreement has been observed between the different methods .
this agreement brings an important support to the escape - rate formalism as a method to establish a connection between the transport properties here of viscosity and the underlying microscopic chaotic dynamics .
the agreement therefore confirms the theoretical results of the escape - rate formalism @xcite .
this confirmation opens perspectives for our understanding of the hydrodynamic modes of shear viscosity .
indeed , in the case of diffusion , the fractal character of the repeller has been shown to imply that the nonequilibrium states are described by singular measures defined in the phase space of the system .
these results suggest a similar scenario in the case of viscosity that the nonequilibrium states of a sheared fluid should also be described by singular measures at the phase - space level of description .
our results can be extended to investigate the connection between viscosity and underlying chaos in many - particle systems with a whole spectrum of positive lyapunov exponents . in many - particle systems ,
the fractal repeller is characterized by a spectrum of partial fractal dimensions which enter the chaos - transport formula ( [ visc - codim2 ] ) . in this way
, we can decompose a transport property such as viscosity onto the spectrum of lyapunov exponents .
the escape - rate formalism also opens perspectives to study viscosity and other transport properties in nanoscopic systems such as nanopipes or atomic and molecular clusters .
indeed , the escape - rate formalism provides a way to define the transport properties already in nanoscopic systems containing a very small number of particles .
the escape - rate formalism is particularly appropriate for nanoscopic systems because the transport coefficients can be defined with the escape rate of the helfand moment out of a finite range of variation . since the helfand moment for viscosity
can be seen as the center of momenta of the particles its variations are limited to a finite range in a system containing a finite number of particles . in this way
, the transport properties could be studied for nanoscopic systems of increasing size in a similar way as the equilibrium thermodynamic properties have been studied in nanoscopic systems as a function of their size .
* acknowledgments . *
the authors thank professors j. r. dorfman and g. nicolis for support and encouragement in this research , and dr .
i. claus for helpful discussions .
sv thanks the fria for financial support .
pg thanks the national fund for scientific research ( fnrs belgium ) for financial support . | we apply the escape - rate formalism to compute the shear viscosity in terms of the chaotic properties of the underlying microscopic dynamics .
a first passage problem is set up for the escape of the helfand moment associated with viscosity out of an interval delimited by absorbing boundaries . at the microscopic level of description ,
the absorbing boundaries generate a fractal repeller .
the fractal dimensions of this repeller are directly related to the shear viscosity and the lyapunov exponent , which allows us to compute its values .
we apply this method to the bunimovich - spohn minimal model of viscosity which is composed of two hard disks in elastic collision on a torus .
these values are in excellent agreement with the values obtained by other methods such as the green - kubo and einstein - helfand formulas
. 0.3 cm ] |
a foundation for the understanding of critical phenomena was provided by wilson s renormalization group ( rg ) approach @xcite which maps the critical point onto a fixed point of a certain transformation of the system s hamiltonian , langevin equation , etc .
the rg theory presents a powerful tool to estimate the values of the critical exponents , it allows to predict which parameters determine the universality class and it explains the existence of an upper critical dimension @xmath0 above which the mean - field theory applies .
furthermore , crossover phenomena between two different universality classes are well understood in terms of competing fixpoints ( see for instance @xcite ) .
nevertheless there are still some open aspects of crossover phenomena which are discussed in the literature .
the question whether the crossover scaling functions are universal was revisited several times @xcite .
for instance it was shown recentlty that two different models , belonging to the same universality class , are characterized by the same ( universal ) crossover scaling functions @xcite .
this is remarkable , since the universal scaling behavior is usually restricted to a small vicinity around the critical point . in the case of a crossover
the universal scaling functions span several decades in temperature or conjugated field .
another question of interest concerns the so - called effective exponents @xcite which can be defined as logarithmic derivatives of the corresponding scaling functions .
it is still open whether these effective exponents fulfill over the full crossover the well - known scaling laws which connect critical exponents .
this question is closely related to the more general and very important question whether effective exponents obey the scaling laws at all .
for instance it is known experimentally @xcite as well as theoretically @xcite that the asymptotic scaling behavior is often masked by corrections to scaling , so - called confluent singularities . in this case
it is useful to analyze the data in terms of effective exponents and the above question naturally arises @xcite .
thus the validity of the scaling laws for effective exponents was addressed in experimental works , rg approaches as well as numerical simulations .
in particular the violation of the scaling laws for effective exponents was conjectured from a rg approximation @xcite .
but neither experimental nor numerical work could clearly confirm this conjecture so far .
for instance binder and luijten considered numerically a crossover in the ising model and discussed the validity of the rushbrook scaling law @xcite .
the observed nonmonotonic crossover behavior suggests again a violation of the rushbrook scaling law .
however it can not be considered as a rigorous proof . in this work
we consider a field driven crossover ( in the so - called critical crossover limit @xcite ) in two different models exhibiting a non - equilibrium second order phase transition .
varying the range of interactions we investigate the crossover from mean - field to non - mean - field scaling behavior .
the order parameter and the order parameter susceptibility is measured as a function of the conjugated field and we are able to determine the corresponding effective exponents over the full crossover regime .
our results show that the well - known widom scaling law is clearly violated for the effective crossover exponents .
furthermore we present a simple analytical argument , suggesting that the scaling laws are valid for the asymptotic scaling regimes ( where the systems are characterized by a pure algebraic behavior ) , whereas the scaling laws do not hold for the crossover regime ( characterized by a non algebraic behavior ) .
in the following we consider two different cellular automata exhibiting a so - called absorbing phase transition . the first model is the conserved transfer threshold process ( cttp ) @xcite . in this model lattice sites
may be empty , occupied by one particle , or occupied by two particles .
empty and single occupied sites are considered as inactive whereas double occupied lattice sites are considered as active . in the latter case one tries to transfer both particles of a given active site to randomly chosen empty or single occupied nearest neighbor sites .
the second model is a modified version of the manna sandpile model @xcite , the fixed - energy manna model @xcite .
in contrast to the cttp the manna model allows unlimited particle occupation of lattice sites .
lattice sites which are occupied by at least two particles are considered as active and all particles are moved to the neighboring sites selected at random . in our simulations
( see @xcite for details ) we have used square lattices ( with periodic boundaries ) of linear size @xmath1 .
all simulations start from a random distribution of particles . after a transient regime
both models reach a steady state characterized by the density of active sites @xmath2 .
the density @xmath2 is the order parameter and the particle density @xmath3 is the control parameter of the absorbing phase transition , i.e. , the order parameter vanishes at the critical density @xmath4 according to , @xmath5 , with the reduced control parameter @xmath6 . below the critical density ( in the absorbing phase )
the order parameter is zero in the steady state .
similar to equilibrium phase transitions it is possible in the case of absorbing phase transitions to apply an external field @xmath7 which is conjugated to the order parameter .
the conjugated field has to act as a spontaneous creation of active particles , destroying the absorbing state and therefore the phase transition itself .
furthermore , the associated linear response function @xmath8 has to diverge at the critical point ( @xmath9 , @xmath10 ) .
a realization of the external field for absorbing phase transitions with a conserved field was recently developed in @xcite where the external field triggers movements of inactive particles which may be activated in this way . at the critical density
@xmath11 the order parameter scales as @xmath12 . using the conjugated field it is possible to investigate the equation of state @xmath13 , i.e. , the order parameter as a function of both the control parameter and the external field .
a recently performed scaling analysis reveals that the cttp and the manna model are characterized by the same critical exponents as well as by the same universal scaling form of the equation of state , i.e. , both models belong to the same universality class @xcite . according to the above definition particles of active sites
are moved to nearest neighbors only , i.e. , the range of interactions is @xmath14 .
it is straightforward to implement various ranges of interactions into these models @xcite . in this
modified models particles of active sites are moved ( according to the rules of each model ) to randomly selected sites within a radius @xmath15 . since the dynamics of both considered models
is characterized by simple particle hopping processes , various interaction ranges can be easily implemented and high accurate data are available . this is a significant advantage compared to e.g. equilibrium system like the ising model where the increasing interaction range causes a slowing down of the dynamics . for any finite interaction
range the phase transition is characterized by non - mean - field scaling behavior which now takes place at the critical density @xmath16 .
a mean - field phase transition occur for infinite interactions ( @xmath17 ) only . but mean - field behavior could occur away from the critical point if the long range interactions reduce the critical fluctuations sufficiently .
the crossover between the mean - field and non - mean - field scaling regimes is described by the famous ginzburg criterion @xcite which states that the mean - field picture is self - consistent in the active phase as long as the fluctuations within a correlation volume are small compared to the order parameter itself .
this leads for zero field to the crossover condition @xmath18 = 1 $ ] , with the crossover exponent @xmath19 @xcite . in order to avoid lattice effects we use the effective range of interactions @xcite @xmath20 where @xmath21 denotes the number of lattice sites within a radius @xmath15
the crossover scaling function has to incorporate three scaling fields ( the control parameter , the external field , and the range of interactions ) , i.e. , we make the phenomenological ansatz @xmath22 where the universal scaling function @xmath23 is the same for all models belonging to a given universality class whereas all non - universal system - dependent features ( e.g. the lattice structure , the update scheme , etc . )
are contained in the so - called non - universal metric factors @xmath24 , @xmath25 , and @xmath26 @xcite .
these factors are determined by three conditions which normalize the scaling function @xmath23 .
first , the analytically known mean - field scaling function @xcite @xmath27 should be recovered for @xmath28 , i.e. , @xmath29 .
therefore , @xmath30 , @xmath31 , implying @xmath32 and @xmath33 .
finally , the non - universal metric factor @xmath26 can be determined by the condition @xmath34 for @xmath35 yielding @xcite @xmath36 the metric factors were already determined in previous works @xcite , thus no parameter fitting is needed in order to perform the following scaling analysis .
of the order parameter of the cttp and manna model at the critical density for @xmath37 .
the metric factors are given by @xmath38 and @xmath39 .
the dashed lines correspond to the asymptotic behavior of the two - dimensional system ( @xmath40 ) and of the mean - field behavior ( @xmath41 ) .
the universal amplitude is given by @xmath42 .
the inset displays the order parameter of the cttp for various values of the interaction range @xmath43 ( from top to bottom ) .
the dashed lines are just to guide the eyes .
, width=302 ] in this work we focus our attention to the field driven crossover , i.e. , we consider the cttp and the manna model at the critical densities @xmath44 which were determined in @xcite . in fig.[fig : uni_co_opcp_03 ] we plot the corresponding data of the cttp for various values of the interaction range @xmath15 . as one can see the power law behavior of the order parameter changes with increasing range of interactions . the scaling form at the critical point is given by ( setting @xmath45 ) @xmath46 with @xmath47 and @xmath48 , respectively . for sufficiently small field
the universal function scales as @xmath49 with the universal amplitude @xmath50 .
the scaling form eq.([eq : scal_plot_opcp_co ] ) has to equal for @xmath14 the @xmath51-dimensional scaling ansatz [ @xmath52 ] leading to @xmath53 according to the scaling form eq.([eq : scal_plot_opcp_co ] ) we plot in fig.[fig : uni_co_opcp_03 ] the rescaled order parameter @xmath54 as a function of the rescaled field @xmath55 .
we observe an excellent data collapse for the full crossover behavior confirming the above phenomenological scaling ansatz .
however , since the entire crossover region covered several decades it could be difficult to observe small but systematic differences between the scaling functions of both models .
therefore , it its customary to scrutinize the crossover via the so - called effective exponent @xcite @xmath56 the perfect collapse of the corresponding data is shown in fig.[fig : uni_co_beta_eff ] and confirms again the universality of the crossover scaling function @xmath23 . .
, width=302 ] next we consider the order parameter susceptibility .
the scaling form of the susceptibility is given by @xmath57 on the other hand the susceptibility is defined as the derivative of the order parameter with respect to the conjugated field @xmath58 with @xmath59 . by comparing this expression with eq.([eq : scal_ansatz_sucp_co ] ) we find @xmath60 , @xmath61 , as well as the widom scaling law @xmath62 which is well known from equilibrium phase transitions .
again , the mean - field behavior is recovered for @xmath28 , i.e. , @xmath63 , implying @xmath64 , @xmath65 , as well as @xmath66 . of the susceptibility of the cttp and the manna model at the critical density for @xmath37 .
the metric factors are given by @xmath38 and @xmath67 .
the dashed lines correspond to the asymptotic behavior of the two - dimensional system ( @xmath68 ) and of the mean - field behavior ( @xmath69 ) .
the universal amplitude is given by @xmath70 .
the inset displays the corresponding effective exponent @xmath71 .
, width=302 ] similar to the order parameter we plot the susceptibility according to the scaling form @xmath72 approaching the transition point the susceptibility is expected to scales as @xmath73 , for @xmath74 , where the universal power - law amplitude is given by @xmath75 the rescaled susceptibility is shown in fig.[fig : uni_co_sucp_02 ] . over the entire crossover region
we got an excellent data collapse including both asymptotic scaling regimes .
the inset displays the effective exponent @xmath76 which exhibits again a monotonic crossover from the two - dimensional scaling regime to the mean - field scaling behavior .
in this way we have obtained the effective exponents @xmath77 and @xmath78 for the field driven crossover from mean - field to non - mean - field behavior .
thus we are able to check the corresponding widom scaling law @xmath79 for the whole crossover region .
the corresponding data are shown in fig.[fig : widom_01 ] . as can be seen the widom scaling law
is fulfilled for the asymptotic regimes ( @xmath37 scaling behavior and mean - field scaling ) but it is clearly violated for the intermediate crossover region . this result is not surprising if one notices that the above widom law [ eq.([eq : widom_eff ] ) ] corresponds to the differential equation [ see eqs.([eq : def_eff_exp_beta_sigma ] , [ eq : def_eff_exp_gamma_sigma ] ) ] @xmath80 using @xmath81 we get @xmath82 where @xmath83 is some constant .
it is straightforward to show that this differential equation is solved by simple power - laws [ @xmath84 with @xmath85 .
thus the widom scaling law is fulfilled in the asymptotic regimes only . in the case that the scaling behavior is affected by crossovers , confluent singularities , etc .
no pure power - laws occur and the scaling laws do not hold for the corresponding effective exponents . ) ] in the crossover regime .
, width=302 ]
in conclusion , the crossover from mean - field to non - mean - field scaling behavior is numerically investigated for two different models exhibiting a second order phase transition .
increasing the range of interactions we are able to cover the full crossover region which spans several decades of the conjugated field .
the corresponding data show that the widom scaling law is violated in the crossover regime .
notice that we focus in our investigations on the particular universality class of absorbing phase transitions only for technical reasons .
the demonstrated violation of the widom scaling can be applied to continuous phase transitions in general . | in this work we consider the universal crossover behavior of two non - equilibrium systems exhibiting a continuous phase transition .
focusing on the field driven crossover from mean - field to non - mean - field scaling behavior we show that the well - known widom scaling law is violated for the effective exponents in the so - called crossover regime . |
one of the most important events expected in particle physics in near future is unquestionably the discovery of the higgs boson . the data from the atlas and the cms experiments at the lhc and the dzero and the cdf experiments at the tevatron hint at the existence of a light standard - model - like higgs boson in the mass range of 115 - 130 gev/@xmath1 @xcite .
if it is indeed the case , the discovery is expected to be declared within a year or so by the lhc experiments . in the standard model ,
the higgs boson is responsible for giving masses to both gauge bosons and matter fermions , via the gauge and yukawa interactions , respectively , upon the spontaneous breaking of the electroweak symmetry .
however , unlike the gauge interaction , the mechanism of the spontaneous symmetry breaking and the yukawa interaction have been left untested . as a matter of fact ,
a higgs doublet with its wine bottle potential , and its yukawa coupling to each matter fermion in the standard model are mere assumptions other than being the minimal mechanism to generate the masses of gauge bosons and fermions . in other words ,
we know essentially nothing but something must be condensed in the vacuum to give the masses of gauge bosons and fermions .
it is well known that the standard model can not describe everything in the universe .
an example is the existence of the dark matter which occupies about one - fourth of the energy density in the universe .
the non - existence of anti - matter is another example .
since the gauge sector of the standard model is well tested , it would be natural to expect that some hints of physics beyond the standard model could be obtained via precise measurements of the higgs boson properties .
the lhc experiments are likely to discover the standard - model - like higgs boson
. however , their precision is most probably not enough to reveal details of the discovered particle(s ) due to high background environments of proton - proton collisions .
thus , precise measurements of the higgs boson properties by an electron - positron collider and its possible options are crucial to uncover its detailed properties which might go beyond the standard model . the international linear collider ( ilc )
has potential to study the properties of the higgs boson(s ) such as coupling strengths to gauge bosons and matter fermions including the top quark with high precision , thereby opening up a window to physics beyond the standard model @xcite .
in addition to the @xmath5 collisions , high energy photon - photon collisions are possible at the ilc by converting the electron beam to a photon beam by the inverse compton scattering @xcite .
physics and technical aspects of a photon linear collider ( plc ) as an option of the @xmath5 linear collider are described , for instance , in ref .
@xcite . a schematic of the plc
is shown in fig .
the higgs boson properties such as its two - photon decay width and cp properties can be studied in high energy photon - photon interaction and thus the plc plays complementary role to the @xmath5 linear collider .
it should also be emphasized that the higgs boson can be singly produced in the @xmath6-channel process so that the required electron beam energy is significantly lower than that for the @xmath5 linear collider .
one of the most important observables to be measured in the higgs sector is its self - coupling , since it directly relates to the dynamics of the higgs potential , _
i.e. _ the mechanism of the spontaneous symmetry breaking .
for example , a non - standard large deviation in the self - coupling can be direct evidence for strong first - order phase transition of the electroweak symmetry in the early universe @xcite .
[ aahheezhh ] ( a ) and ( b ) show diagrams of processes which involve the self - coupling in @xmath0 and @xmath5 interactions . recently , a prospect for studying the self - coupling at the ilc was reported . according to the study ,
the self - coupling is expected to be measured with precision of 57 % with an integrated luminosity of 2 ab@xmath7 at @xmath8 gev @xcite . and ( b ) : for @xmath9 .
higgs boson self - coupling occurs at red points .
] measurements of the self - coupling at the plc were discussed by several authors @xcite .
it has been pointed out that contributions to the self - coupling to the cross - section of fig .
[ aahheezhh ] ( a ) and ( b ) are different and measurements in @xmath5 and @xmath0 interactions are complementary from physics point of view .
in addition , as a higgs boson pair is directly produced in the @xmath0 interaction , required beam energy is lower , 190 gev as described later , than that for the @xmath5 interaction .
this nature is important when considering energy update scenarios of the ilc . in ref .
@xcite , an order - of - magnitude estimation for background processes was presented . however , the cross - section for the @xmath10 boson pair production is @xmath11 orders of higher than that for the higgs boson pair production .
the backgrounds from @xmath3 , @xmath12 , and @xmath13 production processes are also large and have the same final state as with the higgs pairs for a low mass higgs boson , which predominantly decays into @xmath14 .
given the situation , people had been skeptical about the feasibility of the detection of the higgs pair process at the plc . in this work
, we studied , for the first time , the higgs boson pair creation at the plc extensively with a parameter set of the plc based on an @xmath5 linear collider optimized for the light higgs boson of 120 gev/@xmath1 and the same detector simulation framework as used for the ilc physics analysis .
we report details of the analysis , issues and prospects for the measurement of the higgs boson pairs at the plc .
in order to choose parameters for the plc , we calculated the statistical sensitivity @xmath15 defined as @xmath16 where , @xmath17 is the deviation of the self - coupling constant from the standard model . the constant of higgs self - coupling @xmath18 can be expressed as @xmath19 , where @xmath20 is the higgs self - coupling constant in the standard model . @xmath21 and @xmath22 are the expected number of events as a function of @xmath17 and that expected from the standard model . @xmath23 and @xmath24 are the cross - section of the higgs boson production as a function of @xmath17 and that of the standard model , while @xmath25 , @xmath26 , @xmath27 , and @xmath28 are the integrated luminosity , the detection efficiency for the signal , the detection efficiency for backgrounds , and the cross - section of background processes , respectively . for @xmath29 and @xmath30 , @xmath15 is written as @xmath31 fig . [ sensitivity ]
plots @xmath15 as a function of the center of mass energy of the @xmath0 collision ( denoted @xmath32 hereafter ) for the higgs boson mass of 120 gev@xmath33 with the @xmath0 integrated luminosity of 1000 fb@xmath7 .
the cross - section of the signal was calculated according to the formula described in ref .
@xcite for the case of @xmath34 and @xmath35 as indicated in fig .
[ sensitivity ] . from this result
, we found the optimum energy to be @xmath36 270 gev . ) as a function of @xmath0 collision energy .
black and red dots show the @xmath34 and @xmath35 cases . ]
the parameters for the electron and the laser beams are summarized in table [ cain ] .
it was designed to maximize @xmath0 luminosity at @xmath37 gev based on the tesla optimistic parameters @xcite .
the wavelength of the laser was chosen to be 1054 nm , which is a typical wavelength for solid state lasers .
the electron beam energy was chosen to maximize @xmath0 luminosity around 270 gev , while keeping the electron beam emittance and the @xmath38 functions at the interaction point the same as the tesla parameters .
the luminosity distribution was simulated by cain @xcite , as shown in fig .
[ luminosity ] .
the @xmath0 luminosity in the high energy region ( @xmath39 ) was calculated to be @xmath40 .
.the parameters of electron and laser beams based on tesla optimistic parameters .
the polarization of electron beam was assumed to be 100 % . [ cols="^,^,^",options="header " , ] [ cutsummary_cheat ]
we studied the feasibility of the measurement of higgs pair creation at the plc , which is a possible option of the ilc .
the optimum center of mass energy of the @xmath0 collision was found to be around 270 gev for the higgs boson with a mass of 120 gev/@xmath1 .
we found that the @xmath41 process can be observed with a statistical significance of about @xmath42 for the integrated luminosity corresponding to 5 years of the plc running against the background process which has @xmath11 times larger production cross - section ( @xmath43 ) than the signal and other backgrounds which have the same final state ( @xmath44 , and @xmath45 ) , if each track could be successfully assigned to parent particles ( or partons ) .
our analysis showed , for the light higgs boson , improvement of the jet clustering technique is crucial to discriminate the backgrounds by invariant mass information rather than to improve the @xmath46-quark tagging efficiency .
this fact is reasonable , because the @xmath47 background turned out to be suppressed by a simple @xmath46-quark tagging scheme since the @xmath10 bosons do not decay into @xmath46-quark pairs , while the @xmath13 and @xmath3 backgrounds can only be suppressed by their mass differences . for further improvements ,
vertex information from @xmath46-tagging analysis must be taken into account in jet clustering , thus they should be coherently developed .
efforts in this direction are on - going as a part of the ilc physics study @xcite and significant improvement could be expected in near future . the authors would like to thank ilc physics working group @xcite for useful discussions , especially y. okada for useful theoretical inputs
. this work is supported in part by the creative scientific research grant no .
18gs0202 of the japan society for promotions of science ( jsps ) , the jsps core university program , and the jsps grant - in - aid for science research no .
22244031 , and the jsps specially promoted research no .
23000002 .
99 the atlas collaboration , arxiv:1202.1408v1 [ hep - ex ] ( 2012 ) the cms collaboration , arxiv:1202.1488v1 [ hep - ex ] ( 2012 ) http://www.fnal.gov/pub/presspass/press_releases/2012/higgs-boson-20120307.html a. djouadi , j. lykken , k. mnig , y. okada , m. oreglia , s. yamashita _ et al .
_ , `` _ ilc reference design report volume 2 physics at the ilc _ '' ( 2007 ) i. ginzburg , g. kotoki , v. serbo , and v. telnov , pizma ahetf * 34 * 514 ( 1981 ) , jetp lett . *
34 * 491 ( 1982 ) e.boos , a. de roeck , i. f. ginzburg , k. hagiwara , r. d. heuer , g. jikia , j. kwiecinski , d. j. miller , t. takahashi , v. i. telnov , t. rizzo , i. watanabe , p. m. zerwas , arxiv : hep - ph/0103090v1 s. kanemura , y. okada , e. senaha , phys .
b * 606 * 361 ( 2005 ) j. tian , slide at the 23rd general meeting at kek ( 2011 ) r. belusevic , g. jikia , phys .
d * 70 * 073017 ( 2004 ) e. asakawa , d. harada , s. kanemura , y. okada , k. tsumura , phys .
b * 672 * 354 ( 2009 ) e. asakawa , d. harada , s. kanemura , y. okada , k. tsumura , phys .
d * 82 * 115002 ( 2010 ) b.badelek _ et al_. , int .
30 , 5097 ( 2004 ) n. maeda , k. fujii , k. ikematsu , y. kurihara , t. takahashi , proceedings of the ilc physics working group meetings at kek , arxiv:1008.0907 [ hep - ph ] h. murayama , i. watanabe , k. hagiwara , kek report 91 - 11 ( 1992 ) g. j. gounaris , j. layssac , p. i. porfyriadis , f. m. renard , arxiv : hep - ph/9909243v3 t. diakonidis , g. j. gounaris , j. layssac , eur . phys .
j. c * 50 * , 47 - 52 ( 2007 ) f. yuasa _
et al_. , prog .
. suppl . *
138 * ( 2000 ) 18 [ arxiv : hep - ph/0007053 ] s. kawabata , comput .
* 88 * ( 1995 ) 309 acfa linear collider working group , arxiv : hep - ph/0109166v1 http://www-jlc.kek.jp/subg/offl/jsf/ t. sstrand , s. mrenna , p. skands , arxiv : hep - ph/0603175 r. yonamine , k. ikematsu , t. tanabe , k. fujii , y. kiyo , y. sumino , h. yokoya , phys .
d * 84 * 014033 ( 2011 ) k. monig , acta phys .
b * 37 * , no . 4 , 1251 ( 2006 ) w. bartel _
et al_. , z. phys .
c * 33*:23 ( 1986 ) c. peterson , t. rognvaldsson , l. lonnblad , comput .
* 81 * ( 1994 ) 185 , we used a root - base version of jetnet .
http://www.hep.uiuc.edu/home/catutza/root_ to _
jetnet / rtj.html[http://www.hep.uiuc.edu / home / catutza / root _ to _ jetnet / rtj.html ] v. barger , r. phillips , `` _ collider physics _ '' ( addison wesley , 1987 ) t. tanabe , t. suehara , s. yamashita , slide at international workshop on future linear colliders ( lcws11 ) ( 2011 ) t. suehara , t. tanabe , s. yamashita , j. tian , k. fujii , slide at international workshop on future linear colliders ( lcws11 ) ( 2011 ) | we studied the feasibility of the measurement of higgs pair creation at a photon linear collider ( plc ) . from the sensitivity to the anomalous self - coupling of the higgs boson
, the optimum @xmath0 collision energy was found to be around 270 gev for a higgs mass of 120 gev/@xmath1 .
we found that large backgrounds such as @xmath2 and @xmath3 , can be suppressed if correct assignment of tracks to parent partons is achieved and higgs pair events can be observed with a statistical significance of @xmath4 by operating the plc for 5 years . |
nowadays it is well established that the standard model ( sm ) of electroweak interactions describes very well almost every subatomic phenomenon up to order of @xmath4 100 gev @xcite . however , there is a lot of experimental results that are not well understood or nor embedded in their predictions . among the many attempts to generalize the sm ,
several of them involve extensions of the higgs sector .
after spontaneous symmetry breaking , those extensions that add scalar triplet to the higgs sector and left - right symmetry models end up with physical doubly charged higgs bosons ( dchbs ) in its particle spectrum @xcite .
the dchbs ( @xmath5 ) have some interesting consequences .
the main one is related to the seesaw mechanism , which is capable of generating a small mass for the neutrino .
these models are manifested in the higher energy range and the scheme is constructed so that the neutrino mass is inversely proportional to this new energy scale .
this relationship can provide an idea of why the neutrino mass is so small compared with fermi scale @xcite .
one of the major motivations for the study of these particles with double charge comes from the fact that experimental physicists expect to find this kind of particles at the large hadron collider ( lhc ) , ilc and clic .
the fermilab tevatron can detect dchbs via pair production if its mass does not exceed @xmath6 gev an at the lhc this limit rises up to @xmath7 gev . at the tevatron ,
an attempt to directly detect dchbs were carried out based on the channel @xmath8 looking for decay mode @xmath9 , where @xmath10 are known charged leptons .
four - lepton final states was observed by d@xmath11 collaboration @xcite , while in cdf was observed @xmath12 , @xmath13 and states involving @xmath14 leptons @xcite .
these searches were done for left - right model .
then , assuming a br@xmath15 , left ( right ) chiral dchbs with masses larger than 150 ( 136 ) gev and 127 ( 113 ) gev , respectively , are excluded @xcite .
in this paper we work under the 3 - 3 - 1 electroweak model @xcite . in this class of models the electroweak symmetry gauge group su(2)@xmath16@xmath17u(1 ) of the sm
is extended to su(3)@xmath16@xmath17u(1)@xmath18 .
the simplest version of scalar sector of the model has only three higgs triplets @xcite .
consequently , this model predicts the existence of eight physical higgs bosons which are four neutral bosons @xmath19 , @xmath20 , @xmath21 and @xmath22 , a pair of single charged bosons @xmath23 , and a pair of dchbs .
it loses the perturbative character on a scale of a few tev , so that , we can admit that this is the upper limit of energy for the model under consideration @xcite .
this class of models is one of the most interesting alternatives to the sm .
the process of cancellation of anomalies requires that the number of families be an integer multiple of the number of colors .
but , since qcd requires that the number of colors is less than five , then in the 3 - 3 - 1 model the number of families is exactly three @xcite .
we are interested in production of single dchb via interactions @xmath24 from @xmath25 collisions in international linear collider ( ilc ) and in cern linear collider ( clic ) @xcite .
the version of the 3 - 3 - 1 model which we studied predicts the existence of heavy leptons @xmath26
@xmath27 @xcite .
hence , the process to be consider is @xmath28 , which receives contributions of @xmath29 and @xmath30 exchange in the @xmath31 channel and @xmath32 in the @xmath33 channel . in the next section we will present the main features of the version of the 3 - 3 - 1 model under consideration . in sec .
iii we will give the expressions for production cross sections and finally , in sec .
iv we will describe the results and conclusions .
we will summarize in this section only the most relevant points of the model .
for details see ref .
the left - handed leptons and quarks transform under the su(3)@xmath16 gauge group as the triplets -0.4 cm @xmath34 where @xmath35 , @xmath36 , @xmath37 and @xmath38 , @xmath39 .
the @xmath40 exotic quark carries @xmath41 units of elementary electric charge , while @xmath42 and @xmath43 carry @xmath44 each one . in eqs .
( [ fer ] ) the numbers @xmath11 , @xmath45 , and @xmath46 are the u(1)@xmath47 charges .
each left - handed charged fermion has its right - handed counterpart transforming as a singlet in the presence of the su(3)@xmath16 group , i.e. , @xmath48 [ ferr ] we are defining @xmath49 , @xmath50 , @xmath31 and @xmath51 , @xmath33 , @xmath52 .
it should be noted that , in order to avoid anomalies , one of the quark families must transform in a different way with respect to the others . in eqs .
( [ ferr ] ) all the primed fields are linear combinations of the mass eigenstates .
the charge operator is defined by @xmath53 where @xmath54 and @xmath55 are the diagonal gell - mann matrices .
we notice , however , that since @xmath56 in eqs .
( [ fer ] ) are in antitriplet representation of su(3)@xmath16 , the antitriplet representation of the gell - mann matrices must also be used in eq .
( [ op ] ) in order to get the correct electric charge for the quarks of the second and third generations .
the three higgs scalar triplets @xmath57 are the minimal content of higgs sector enough to break the symmetry spontaneously and generate the masses of fermions and gauge bosons in the model .
the neutral scalar fields develop the vacuum expectation values ( vevs ) @xmath58 , @xmath59 and @xmath60 , with @xmath61 gev@xmath62 .
the pattern of symmetry breaking is @xmath63 and so , we can expect @xmath64 then , after the symmetry breaking , the masses of scalar fields become , @xmath65 , \qquad m_{h_1^\pm}^2 = \frac{v_w^2}{2v_\eta v_\rho}\left(fv_\chi - 2\lambda_7v_\eta v_\rho\right ) , } \\ & & { m_{h_2^\pm}^2 = \frac{v_\eta^2 + v_\chi^2}{2v_\eta v_\chi}\left(fv_\rho - 2\lambda_8v_\eta v_\chi\right ) , \qquad m_{h^{\pm\pm}}^2 = \frac{v_\rho^2 + v_\chi^2}{2v_\rho v_\chi}\left(fv_\eta - 2\lambda_9v_\rho v_\chi\right)}.\end{aligned}\ ] ] where in eqs .
( [ mapp ] ) we have considered eq .
( [ app ] ) . due to the transformation properties of the fermion and the higgs fields under su(3)@xmath16 [ see eqs .
( [ ferr ] ) and ( [ higgs ] ) ] the yukawa interactions in the model are @xmath66 + \cr & & + \sum_{\alpha , \beta}f^j_{\alpha\beta}\overline{q}_{\alpha l}j^\prime_{\beta r}\chi^ * + g^j\overline{q}_{1l}j_{1r}\chi + { \rm h. c. } , \label{lq}\end{aligned}\ ] ] [ yuk ] where @xmath67 s , @xmath68 s , @xmath69 s , and @xmath70 s are yukawa coupling constants with @xmath71 , @xmath52 @xmath72 @xmath73 , @xmath74 , @xmath39 and @xmath75 , @xmath76
@xmath72 @xmath74 , @xmath39 .
the interaction eigenstates , which appear in eqs .
( [ yuk ] ) , can be transformed into the corresponding physical eigenstates by appropriated rotations , _
i. e. _ , @xmath77 however , since the cross - section calculations imply summation on flavors ( sec .
iii ) and the rotation matrix must be unitary , the mixing parameters have no essential effects for our purpose here .
so , hereafter we suppress the prime notation for the interaction eigenstates .
the lagrangian ( [ ll ] ) shows that the ordinary particle masses are proportional to @xmath78 and @xmath79 , while the heavy ones are proportional to @xmath80 .
we are not considering here neutrino masses since they are not important in the discussion .
the gauge bosons consist of an octet @xmath81 @xmath82 , associated with su(3)@xmath16 and a singlet @xmath83 associated with u(1)@xmath47 .
the covariant derivative is @xmath84 where @xmath85 @xmath72 @xmath86 , @xmath87 , @xmath88 .
the model predicts single - charged @xmath89 , doubly - charged @xmath90 vector bileptons and a new neutral gauge boson @xmath91 in addition to the charged standard gauge bosons @xmath92 and the neutral standard @xmath93 .
the masses of the new gauge bosons are @xmath94 where @xmath95 .
the charged currents are read off from @xmath96 where @xmath97 is a mixing matrix .
the main motivation of this work is to show that in the context of the 3 - 3 - 1 model of ref .
@xcite , the signatures for dchbs can be significant at the ilc and at the clic .
one way to search for @xmath5 and @xmath98 is through the process @xmath99 , where the photons comes either from bremsstrahlung or from laser backscattering .
our results indicate a satisfactory number of events to establish the signal . analyzing it we can make inferences about the existence of dchbs , doubly charged gauge bosons and heavy leptons .
we study the direct dchbs production in the process @xmath0 , which occurs through the contribution of the @xmath33 channel , as shown in fig .
@xmath73 , and the contributions of the @xmath100 and @xmath101 in the @xmath31 channel ( see fig .
@xmath74 ) .
we assume the intermediate photon is produced either by the weizs@xmath3cker - williams @xcite or by backscattering of laser from the @xmath32 beam @xcite .
the interactions lagrangian is given sec .
[ sec2 ] and in several papers ( see , for example , ref .
then we evaluate the differential subprocess cross section for this reaction as @xmath102.\end{aligned}\ ] ] it should be mention that the interference terms give not any contributions to this cross section and the @xmath103 are given in the form @xmath104 @xmath75 is the fine structure constant which we take equal to @xmath105 , @xmath106 is the center of mass ( cm ) energy of the @xmath24 system and the kinematic invariants are @xmath107 with @xmath108^{1/2}\right\},\end{aligned}\ ] ] where @xmath109 is the angle between the dchb and the incident electron in the cm - frame , in this frame @xmath110 is the vertex strength of the @xmath5 to @xmath111 and @xmath5 ( @xmath112 ) with @xmath113 and @xmath114 being the momentum four - vectors of the @xmath111 and @xmath5 , respectively .
@xmath115 is the vertex strength of the @xmath5 to @xmath116 and @xmath98 , @xmath117 is for @xmath118 to @xmath111 and @xmath5 ( @xmath119 ) , @xmath120 for @xmath118 to @xmath116 and @xmath98 and the others couplings are given in @xcite .
it is to note that in previous papers @xcite , was made an error in the calculation of the coupling @xmath121 , where the index @xmath71 denotes the flavor , this error was now corrected ( @xmath122 ) .
we are assuming @xmath123 @xcite .
so we obtain the total cross section for this process folding @xmath124 with the one photon luminosities @xmath125 where @xmath126 .
here we present the cross section for the process @xmath127 for the ilc ( @xmath128 tev and clic ( @xmath39 tev ) .
all calculations were done according to @xcite and we obtain for the parameters and the vev , the following representative values : @xmath129 , @xmath130 , @xmath131 @xmath132 , @xmath133 , @xmath134 , @xmath135 gev , these parameters and vev are used to estimate the values for the particle masses which are given in table i. referring to the backscattered photons ( fig .
12 ) , considering @xmath136 gev and @xmath137 gev , we will have a total of @xmath138 events of @xmath139 produced per year .
regarding @xmath140 gev and for the same masses above it will give a total of @xmath141 events per year to produce @xmath139 . taking the @xmath136 gev and @xmath137 gev and taking the same first signal as above , that is , br@xmath142 ( fig .
7 ) we will have a total of @xmath143 events per year .
with respect to @xmath140 gev , and @xmath136 gev , and considering now br@xmath144 ( fig . 8) it will give a total of @xmath145 events per year .
considering the second signal for @xmath146 gev and @xmath137 gev and taking into account that br@xmath147 ( fig .
7 ) , and br@xmath148 ( fig . 9 and @xcite ) , we would have @xmath149 events per year . considering @xmath140 gev it will not given any event for @xmath150 for the same reasons given above . for @xmath151
it presents @xmath152 events per year regarding the branching ratios br@xmath153 ( fig .
7 ) and br@xmath154 ( fig 10 and @xcite ) .
all these results are resumed in table vi and table vii . in this way we have as a signal for the process
@xmath155 , this final state can be considered as the golden mode of decay .
the background wich can appear is @xmath156 , but these backgrounds can be readely reduced by impossing the z window cut where the invariant mass of opposite - sign lepton pairs must be far from the z mass : @xmath157 gev , this removes events where the leptons come from the z decay @xcite . these two charged leptons @xmath158 are relatively easy to measure in the detector , as they leave tracks and do not shower like gluons and quarks , by computing the invariant masses of both the pairs . on the other side the signal of a process @xmath159 and @xmath160 ,
are @xmath161 , where the @xmath111 and @xmath162 will be measured with high accuracy by the detector and a sharp doubly charged gauge bosons will be observed in the two same - sign lepton invariant mass distribution .
if we see this signal we will not only be seeing the dchbs but also the doubly charged gauge bosons and heavy leptons .
so the @xmath163 collisions can be also a plentiful source of dchbs .
the discovery of dchbs will be without any doubt of great importance for the physics beyond the sm because of the confirmation of the higgs triplet representation and the verification of the existence of some exotic particles . in summary , through this work
, we have shown that in the context of the 3 - 3 - 1 model the signatures for dchbs can be significant for @xmath163 collisions obtained by bremsstrahlung and backward compton scattering of laser .
this kind of new particles have distinct experimental signals of like - sign dileptons , offering excellent potential for dchbs discovery .
their observation in future high energy collider experiments would be a clear evidence of new physics .
thus , searching for @xmath5 is one of the main goals of current and future high energy collider experiments .
our study indicates the possibility of obtaining a clear signal of these new particles with a satisfactory number of events .
one of us ( j. e. c. m. ) would like to thank to prof .
o. j. p. boli for the proposal of this work and to hospitality of the departamento de fsica matemtica -usp - brazil , where part of this work was done and m. d. t. is beholden to instituto de fsica terica of the unesp for his hospitality and to conselho nacional de desenvolvimento cientfico e tecnolgico for partial support .
this work was supported by fundao de amparo pesquisa no estado de so paulo ( processo no . 2009/02272 - 2 ) .
for example , g. b. gelmini and m. roncadelli , phys .
b * 99 * , 411 ( 1981 ) ; r. n. mohapatra and r. e. marshak , phys . lett . * 91 * , 202 ( 1980 ) ; r. n. mohapatra and and d. sidhu , phys .
* 38 * , 667 ( 1977 ) .
r. n. mohapatra and g. senjanovich , phys .
44 * , 912 ( 1980 ) ; m. gell - mann , p. ramond and r. slansky , in _ supergravity _ , edited by p. van niewenhuizen and d. z. friedman ( north - holland , amsterdam , 1979 ) ; t. yanagida , in _ proceedings of workshop on unified theories and baryon number in the universe _ , edited by o. swada and a. sugamoto ( kek , tsukuba , 1979 ) .
chong - xing yue , xue - song su , jiao zhang , jue wang , arxiv:1010.4633 [ hep - ph ] ; s. chakrabarti , d. choudhury , r. m. godbole , b. mukhopadhyaya , phys .
b * 434 * , 347 ( 1998 ) ; e. m. gregores , a. gusso , s. f. novaes , phys .
d * 64 * , 015004 ( 2001 ) ; s. godfrey , p. kalyniak , n. romanenko , phys .
d * 65 * , 033009(2002 ) ; chong - xing yue , shuang zhao , wei ma , nucl . phys .
b * 784 * , 36 ( 2007 ) . | we studied the production and signatures of doubly charged higgs bosons in the process @xmath0 , where @xmath1 is a heavy lepton , at the @xmath2 international linear collider ( ilc ) and cern linear collider ( clic ) .
the intermediate photons are given by the weizs@xmath3cker - williams and laser backscattering distributions .
we found that significant signatures are obtained by bremsstrahlung and backward comptom scattering of laser .
a clear signal can be obtained for doubly charged higgs bosons , doubly charged gauge bosons and heavy leptons . |
the interest of the astronomical community has been recently caught by a few very luminous supernovae ( sne ) that have exploded within dense , massive and pre - existent circumstellar media .
some of them have been spectroscopically classified as type iin supernovae ( sne iin , * ? ? ?
* ) because of the presence of prominent narrow ( about 1000 km s@xmath0 ) h balmer emission lines ( e.g. sn 2006gy , @xcite ; sn 2007bw , @xcite ) . for few other events ( e.g. sn 1999cq , @xcite ;
sn 2006jc , @xcite ) , whose spectra show relatively narrow he i emission lines ( @xmath1 2000 km s@xmath0 ) , a different nomenclature was coined : type ibn sne @xcite . in the spectra of these objects ,
the narrow h lines are much weaker than the he i features , or completely missing .
in addition , a number of regular type iin sne have developed at some stages clear evidence of narrow ( from a few @xmath2100 km s@xmath0 to @xmath12000 km s@xmath0 ) he i lines although , contrary to sne ibn , they are always less prominent than h@xmath3 .
this group includes sn 1978k @xcite , sn 1987b @xcite , sn 1988z @xcite , sn 1995 g @xcite , sn 1995n @xcite , sn 1996l @xcite , sn 1997ab @xcite , sn 1997eg @xcite , sn 1998s @xcite . in all these sn events ,
the narrow lines ( both of h and he i ) are thought to originate in a slowly - moving circumstellar medium ( csm ) which is photoionized either by early uv flux or interaction with the sn ejecta .
a number of authors @xcite suggest that a significant fraction of sne iin might be produced by the explosion of massive luminous blue variable ( lbv ) stars , even if the nature of at least some of type iin events is still debated ( e.g. sn 2002ic and similar events , see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and references therein ) .
as an alternative view , @xcite propose a pair - instability pulsation scenario as a trigger of major mass ejection episodes .
the ejecta collide with the material expelled by the star in previous eruptions converting large amounts of kinetic energy into radiation and generating a type iin sn .
type ibn sne seem to be connected with wr progenitors whose he envelope is almost completely stripped away , and that explode while they are still embedded within the he - rich material @xcite .
@xcite showed that sn 2006jc was preceded by a giant outburst 2 years before core - collapse .
although the sn itself suggests a wc star progenitor , the outburst had similar characteristics to an lbv eruption . in ( * ? ? ?
* paper 1 of this series ) the common properties of objects similar to sn 2006jc ( sne ibn ) are discussed . in paper
3 @xcite we investigate the origin of the strong infrared excess observed in sn 2006jc . in this paper ( paper 2 )
we describe the observed characteristics of the unusual sn 2005la , an object whose spectra show simultaneously narrow h and he i features , with the most prominent he i lines having comparable strength ( at least at early stages ) as h@xmath3 , and showing unusual evolution with time .
we propose that sn 2005la is a transitional object lying somewhere between the newly defined sne ibn and more classical type iin events .
, width=321 ] .sequence of local standards in the field of kug 1249 + 278 .
the errors reported in brackets are the r.m.s . of the derived average magnitudes . [
cols="^,^,^,^,^",options="header " , ] @xmath4 2170@xmath5210 ; @xmath6 1960@xmath5310 ; @xmath7 3210@xmath5250 . the very rapid evolution of sn 2005la and its moderate apparent luminosity prevented us from performing extensive spectroscopic monitoring , and only 4 spectra were collected .
the first spectrum was obtained with the 9.2-m hobby - eberly telescope ( het ) of the mcdonald observatory ( university of texas , us ) , equipped with the marcario low - resolution spectrograph ( lrs ) on dec .
additionally , 3 spectra were obtained later using the 1.82-m copernico telescope with afosc .
more information on the 4 spectra is reported in tab .
[ tab3 ] , while the complete spectral sequence is shown in fig .
[ fig3 ] . in fig .
[ fig3 ] the spectra of sn 2005la are also compared with a spectrum of the type ibn sn 2006jc at phase + 70@xmath8 @xcite and a very late spectrum of the type iin sn 1995 g @xcite .
this comparison highlights a rough similarity between the spectra of sn 2005la ( especially at 20 - 21@xmath8 after the discovery ) and that of sn 2006jc .
however , h@xmath3 is prominent in sn 2005la ( although much weaker than in sn 1995 g ) .
its h@xmath3 is comparable with the most prominent he i lines ( i.e. @xmath95876 and @xmath97065 ) at early epochs and is stronger than the he i lines only in the last spectrum . fig .
[ fig3b ] shows the evolution of the profiles of h@xmath10 ( + he i @xmath94921 ) , he i @xmath95876 and h@xmath3 at three selected epochs .
the relative strengths of the h and he i lines may change as a result of abundance and , possibly , ionization effects .
the latter is mostly expected in objects whose ejecta are interacting with a csm . unlike sn 2005la , the spectrum of sn 2006jc shows a very weak ( practically invisible ) h@xmath3 during the first 3 months @xcite .
this line becomes more prominent only at very late epochs @xcite .
instead , in type iin sne , a narrow ( typically with v@xmath11 1000 km s@xmath0 ) h@xmath3 dominates in strength at any phase over the he i lines . in tab .
[ tab4 ] we report the velocities of sn 2005la as derived from measuring the full - width - half - maximum ( fwhm ) of a number of lines ( h and he i ) .
while the he i lines show a clear p - cygni profile , this is not obvious for the h lines .
the velocities derived from the position of the p - cygni minima of the he i lines are quite consistent ( @xmath1 2000 - 2200 km s@xmath0 , see also tab .
[ tab4 ] ) with the fwhm velocities ( v@xmath12 ) of the same lines .
we find that in the 7th december spectrum the v@xmath12 derived for the h lines is about 1500 km s@xmath0 , while it is slightly higher for the he i lines ( v@xmath12 @xmath13 2150 km s@xmath0 ) .
these lines are probably produced in an unshocked csm surrounding the sn .
once the sn ejecta reach the csm , this material is expected to accelerate .
a bump near 7800 is tentatively identified as a proper sn ejecta feature , probably o i @xmath97774 .
the low s / n does not allow us to properly measure the velocity of this feature , but ( despite the uncertainty ) it is significantly higher ( v@xmath12 = 4900@xmath51200 ) than the other lines identified in the december 7th spectrum .
this velocity can be considered as representative for the ejecta velocity .
, width=325 ] the two subsequent spectra ( december 10 and 11 ) show that the line velocities are increasing ( v@xmath12 @xmath13 1900 km s@xmath0 for the h lines ; v@xmath12 @xmath13 3150 km s@xmath0 for the he lines ) , indicating an evident acceleration of the he- and , more marginally , of the h - rich cs region . in the last spectrum of sn 2005la , obtained about two weeks later , all lines are in pure emission ( see fig . [ fig3b ] ) , as expected when driven by the high - energy radiation resulting from the interaction of the ejecta with the dense csm close to the sn . moreover , the velocity of both the h and he i lines is much higher , being v@xmath12 @xmath13 4200 km s@xmath0 , which is close to that estimated for the o i @xmath97774 line .
no extended wings are observed for the emission features , as one would expect if the line broadening was a result of scattering of photons of the narrow lines by the thermal electrons of the circumstellar gas ( as was observed in sn 1998s , * ? ? ?
it is remarkable that the absorption features in the region between 4600 and 5200 ( possibly due to fe ii lines and detected with comparable widths also in sne ib / c ) , the overall blue colour and the drop of the flux beyond 5400 , seem to be common characteristics of most interacting core - collapse sne ( fig .
[ fig3 ] ) .
sn 2005la shows clear evidence of the presence of csm produced by the precursor star before the sn explosion .
the sn spectra indicate that this material is rich in both h and he .
however there are some spectroscopic indications that the inner csm is more he - rich , while h is more abundant in more external regions of the csm . from the available data ,
it is very difficult to understand if this material was produced through a dense , continuum wind or via major outburst episodes similar to those observed in a number of lbvs ( or , alternatively , in luminous outbursts of wr stars analogous to the transient ugc 4904-v1 which preceded by two years
the explosion of sn 2006jc , see * ? ? ?
* ) . up to now
, it was commonly accepted that core - collapse explosions occur substantially ( @xmath1 a few times 10@xmath14 years ) after the last major mass - loss episode
. however , the precedent of sn 2006jc should give impulse to the efforts of finding evidence for similar luminous outbursts occurring only a few years before the explosion of interacting core - collapse sne .
unfortunately , the host galaxy of sn 2005la was not frequently monitored in the past , and very few images are currently available in public archives .
we analysed b and sloan r band images of kug 1249 + 278 ( with exposure times of 3@xmath2300s and 600s , respectively ) obtained on 2002 march 10th with the 2.5-m isaac newton telescope in la palma ( canary islands , spain ) .
the data have been downloaded from the ing archive .
another set of images were obtained making use of the smoka science archive .
the host of sn 2005la was observed at the 1.05-m schmidt telescope of the kiso observatory ( university of tokyo , japan ) on 2003 march 11th ( 4@xmath2300s and 3@xmath2300s in the v and r bands , respectively ) .
finally , the texas sn search has intensively ( almost daily ) monitored the region of kug 1249 + 278 ( from november 20 , 2004 to july 7 , 2005 ) , but no evidence of pre - sn outbursts was found ( fig . [ fig4 ] ) .
this is not really unexpected since the deepest magnitude limits registered for the rotse - iiib telescopes are near magnitude 20 . with the assumptions on the distance and extinction discussed in sect .
[ general ] , only outbursts more luminous than m@xmath15 @xmath1 -14.5 would be detected , which is an absolute magnitude very close to the brightest eruptions reported in fig .
[ fig4 ] , and only @xmath12.7 magnitudes fainter than the peak magnitude registered for sn 2005la .
therefore , the analysis of the pre - explosion images do not provide any useful information on the mass loss processes that occurred in the last years of life of the progenitor of sn 2005la .
, width=326 ]
the spectroscopic properties of sn 2005la , especially the peculiar intensity ratio between h@xmath3 and the most prominent he i lines ( @xmath95876 and @xmath97065 ) suggest that this object could be classified as lying somewhere between the classical type iin sne @xcite and the type ibn events @xcite . however , the rapidly evolving light curve of sn 2005la , and the very fast decline of the optical luminosity ( see fig . [ fig2 ] ) is more typical for the family of sn 2006jc .
there now appears to be strong observational evidence that normal type ii - p sne have red supergiant progenitors which have fairly moderate initial masses @xcite .
it is interesting to note that there are no discoveries of any red supergiants with masses definitely more than 20m@xmath16 .
the case of the blue supergiant exploding as sn 1987a is still not well understood from the stellar evolutionary point of view , but binarity is the most likely explanation @xcite . hence one might conjecture that higher mass stars explode as some other type of event . there is a reasonable probability that wolf - rayet ( wr ) stars of type wn and wc / o are the direct progenitors of at least some sne ib and ic .
recent modelling of the light curves of type ic sne have yielded ejecta masses ( c and o dominated ) that have some overlap with the core masses of wc stars determined from luminosity measurements and stellar evolutionary models . for example , ejecta masses of 2.5 - 5m@xmath16 for sn 2002ap @xcite and 3.5 - 8m@xmath16 for sn 2004aw @xcite were determined , while suggest masses of 7 - 19m@xmath16 for wc stars in the large magellanic cloud ( lmc ) and the galaxy .
while these mass ranges just about overlap , the fact that the explosion of sn 2002ap was asymmetric and lack of a large sample of sn ib / c ejecta or progenitor mass estimates still leaves open the question of the origins of these events .
there is also mounting evidence that type ib / c sne are associated with environments which show stronger h@xmath3 emission and have higher surface brightness than do the environments of type ii - p sne .
this would suggest the progenitors are more likely from higher mass stars , and hence above 15 - 20m@xmath16 .
although rather unexpectedly the spatial type ib sn distribution of @xcite resembles the type ii sn distribution .
there is one possible detection of a type iin sn progenitor , that of sn 2005gl @xcite , although the object is exceptionally bright for a single star and it remains uncertain if it is actually one of the most luminous stars that can exist , a very high - mass lbv or if it is actually an unresolved stellar cluster . in the case of the type iin sn 2005gl ,
the ejecta started to interact early with the material likely lost through major eruptions a short time before the explosion of the star . from this point of view
, there is a remarkable analogy with the sequence of events that led to sn 2006jc in ugc 4904 , as a luminous pre - sn outburst followed two years later by the real sn explosion @xcite .
in such a case , if the precursor of sn 2006jc was a single massive star , it was responsible for the outburst observed on october 2004 @xcite , in which a significant amount of material was likely expelled .
sn 2006jc was monitored in detail for several months after october 2006 , and the spectra showed relatively narrow he i emission lines , indicating that the sn ejecta were interacting with a slower moving he - rich csm . as a consequence ,
the 2004 luminous outburst was probably produced by a wr star transiting between the wn and wc / o phases ( e.g. * ? ? ?
* ; * ? ? ?
perhaps _ all _ precursor stars of the interacting type ibn sne are rare objects @xcite which lie in the region between more typical wn and wc / o stars ( which should produce sne ib and sne ic , respectively ) .
one might draw a parallel between this argument and that used to relate the lbv progenitors of sne iin to the rsg / bsg progenitors of non - interacting sne ii . in this context ,
sn 2005la is unique , because it shares properties with both sne iin and sne ibn .
hence we suggest that sn 2005la might have been produced by the explosion of a star which was transiting from the lbv to the wr ( wn - type ) phase . from the analysis of the sn properties , the young wn precursor of sn 2005la should have suffered a major mass loss episode ( or possibly several episodes ) a few years before the sn explosion .
one would expect that during these episodes a significant amount of he - rich material ( with residual h ) was lost .
the detection of outbursts similar to that seen for sn 2006jc and other similar sne would be an exciting discovery as it would hint that the instability may be directly related to the imminent core - collapse .
unfortunately , the analysis of the available pre - sn archive images for sn 2005la ( see sect . [ outburst ] ) does not provide any direct confirmation for this scenario although , of course , it does not exclude it .
wide field transient searches which sample low - mass and hence low metallicity galaxies are the obvious way forward to discover more of these peculiar and interesting sne .
for example the blank - field strategy of the texas supernova search has discovered a large number of peculiar sne ( e.g. the weird sn 2005ap , * ? ? ?
* ) . the future pan - starrs project @xcite will have the capability of increasing these discoveries further and more extensive monitoring of the events will help to shed light on the post - main - sequence evolution of the most massive stars .
sm acknowledges financial support from the academy of finland ( project : 8120503 ) . + the paper is based on observations collected at the 0.45 m rotse - iiib robotic telescope and the hobby - eberly telescope at the mcdonald observatory ( university of texas , us ) , the 0.45 m rotse - iiid robotic telescope of the turkish national observatory ( bakirlitepe , turkey ) , the copernico 1.82-m telescope of the inaf - asiago observatory ( mt . ekar , italy ) and the liverpool telescope ( la palma , spain ) .
we thank the rotse collaboration , the hobby - eberly telescope staff , particularly b. roman , v. riley , and s. rostopchin , and the support astronomers of the liverpool telescope for performing the observations of sn 2005la . +
this paper makes use of data obtained from the isaac newton group archive which is maintained as part of the casu astronomical data centre at the institute of astronomy , cambridge .
this work is also based in part on data collected at the kiso observatory ( university of tokyo ) and obtained from the smoka science archive , which is operated by the astronomy data center , national astronomical observatory of japan .
this research 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 .
we acknowledge the usage of the hyperleda database ( http://leda.univ-lyon1.fr ) .
+ 99 antilogus , p. et al .
2007 , cbet 941 anupama , g. c. , sivarani , t. , pandey , g. 2001 , a&a , 367 , 506 aretxaga , i. , benetti , s. , terlevich , r. j. , fabian , a. c. , cappellaro , e. , turatto , m. , della valle , m. 1999 , mnras , 309 , 343 arkharov , a. , efimova , n. , leoni , r. , di paola , a. , di carlo , e. , dolci , m. 2006 , astron .
961 baba , h. et al .
2002 , adass xi , eds .
d. a. bohlender , d. durand @xmath17 t. h. handley , asp conference series , vol .
281 , 298 benetti , s. , turatto , m. , cappellaro , e. , danziger , i. j. , mazzali , p. a. 1999 , mnras , 305 , 811 benetti , s. , cappellaro , e. , turatto , m. , taubenberger , s. , harutyunyan , a. , valenti , s. 2006 , apj , 653l , 129 chugai , n. n. , danziger , i. j. , della valle , m. 1995 , mnras , 276 , 530 chugai , n. n. 2001 , mnras , 326 , 1448 chugai , n. n. et al .
2004 , mnras , 352 , 1213 crowther , p. a. , dessart , l. , hillier , d. j. , abbott , j. b. , fullerton , a. w. 2002 , a&a , 392 , 653 di carlo , e. et al .
2008 , apj in press ( arxiv:0712.3855 ) dopita , m. a. , cohen , m. , schwartz , r. d. , evans , r. 1984 , apj , 287l , 69 ercolano , b. , barlow , m. j. , sugerman , b. e. k. 2007 , mnras , 375 , 753 fassia , a. et al .
2001 mnras , 325 , 907 filippenko , a. v. , foley , r. j. , matheson , t. 2005 , iau circ .
8639 , 2 foley , r. j. , smith , n. , ganeshalingam , m. , li , w. , chornock , r. , filippenko , a. v. 2007 , apj , 657l , 105 fransson , c. et al .
2002 , apj , 572 , 350 gal - yam , a. et al .
2007 , apj , 656 , 372 gerardy , c. l. , fesen , r. a. , hflich , p. , wheeler , j. c. 2000 , aj , 119 , 2968 gerardy , c. l. et al .
2002 , apj , 575 , 1007 hamuy , m. et al .
2003 , nature , 424 , 651 hagen , h .- j . , engels , d. , reimers , d. 1997 , a&a , 324l , 29 immler , s. et al .
2007 , apj , 674l , 85 itagaki , k. et al .
2006 , iau circ .
8762 james , p. a. , anderson , j. p. 2006 ,
a&a , 453 , 57 kelly , p. l. , kirshner , r. p. , pahre , m. 2007 , apj submitted ( arxiv:0712.0430 ) kotak , r. , vink , j. s. 2006 , a&a , 460l , 5 landolt , a. u. 1992 aj , 104,340 leonard , d. c. , filippenko , a. v. , barth , a. j. , matheson , t. 2000 apj , 536 , 239l li , w. , wang , x. , van dyk , s. d. , cuillandre , j .- c . , foley , r. j. , filippenko , a. v. 2007 , apj , 661 , 1013 liu , q .- z . , hu , j .- y . , hang , h .-
r . , qiu , y .-
l . , zhu , z .- x . , qiao , q .- y . 2000 ,
a&as , 114 , 219l matheson , t. et al .
2000 , aj , 119 , 2303 mattila , s. et al . 2008 ,
mnras accepted ( arxiv:0803.2145 ) mazzali , p. a. et al . , 2002 ,
apj , 572l , 61 meikle , w. p. s , et al .
2007 , apj , 665 , 608 ofek , e. o. et al .
apj , 659l , 130 pastorello et al .
2002 , mnras , 333,27 pastorello , a. , aretxaga , i. , zampieri , l. , mucciarelli , p. , benetti , s. 2005 , _ 1604 - 2004 : supernovae as cosmological lighthouses _ , asp conference series , vol .
342 , proceedings of the conference held 15 - 19 june , 2004 in padua , italy .
edited by m. turatto , s. benetti , l. zampieri , and w. shea .
san francisco : astronomical society of the pacific , 2005 , p.285 pastorello , a. et al . 2007a , nature , 447 , 829 pastorello , a. et al .
2007b , mnras , 376 , 1301 pastorello , a. et al .
2008 , mnras , accepted ( arxiv:0801.2277 ) prieto , j. l. et al .
2007 , aj submitted ( arxiv:0706.4088 ) prieto , j. l. , stanek , k. z. , beacom , j. f. 2008 , apj , 673 , 999 puckett , t. , sostero , g. , quimby , r. , mondol , p. 2005 , iau circ .
8639 podsiadlowski , p. 1992 , pasp , 104 , 717 pozzo , m. , meikle , w. p. s. , fassia , a. , geballe , t. , lundqvist , p. , chugai , n. n. , sollerman , j. 2004 , mnras , 352 , 457 quimby , r. m. 2006 , ph.d .
thesis , univ .
texas quimby , r. m. , wheeler , j. c. , hflich , p. , akerlov , c. w. , brown , p. j , rykoff , e. s. 2007a , apj , 666 , 1093 quimby , r. m. , aldering , g. , wheeler , j. c. , hflich , p. , akerlov , c. w. , rykoff , e. s. 2007b , apj , 668l , 99 sakon , i. et al .
2007 , apj submitted ( arxiv:0711.4801 ) salamanca , i. , terlevich , r. j. , tenorio - tagle , g. 2002 , mnras , 330 , 844 sollerman , j. , cumming , r. j. , lundqvist , p. 1998
, apj , 493 , 933 schlegel , d. j. , finkbeiner , d. p. , davis , m. 1998 , apj , 500 , 525 schlegel , e. m. 1990 , mnras , 244 , 269 schlegel , e. m. , kirshner , r. p. , huchra , j. p. , schild , r. e. 1996 , aj , 111 , 2038 smartt , s. j. et al .
2004 , science , 303 , 499 smith , n. et al .
2007 , apj , 666 , 1116 smith , n. , foley , r.j .
, filippenko , a. v. 2008 , apj , 680 , 568 stathakis , r. a. , sadler , e. m. 1991 , mnras , 250 , 786 sugerman , b. e. k. et al .
2006 , science , 313 , 196 taubenberger , s. et al .
2006 , mnras , 371 , 1459 tominaga , n. et al .
2007 , apj submitted ( arxiv:0711.4782 ) tremonti , c. a. et al .
2004 , apj , 613 , 898 trundle , c. , kotak , r. , vink , j. s. , meikle , w. p. s. 2008 , a&a , 483l , 47 turatto , m. , cappellaro , e. , danziger , i. j. , benetti , s. , gouiffes , c. , della valle , m. 1993 mnras , 262 , 128 young d. r. et al .
2008 , a&a submitted zampieri , l. , mucciarelli , p. , pastorello , a. , turatto , m. , cappellaro , e. , benetti , s. 2005 , mnras , 364 , 1419 woosley , s. e. , blinnikov , s. , heger , a. 2007 , nature , 450 , 390 | we present photometric and spectroscopic data of the peculiar sn 2005la , an object which shows an optical light curve with some luminosity fluctuations and spectra with comparably strong narrow hydrogen and helium lines , probably of circumstellar nature .
the increasing full - width - half - maximum velocity of these lines is indicative of an acceleration of the circumstellar material .
sn 2005la exhibits hybrid properties , sharing some similarities with both type iin supernovae and 2006jc - like ( type ibn ) events .
we propose that the progenitor of sn 2005la was a very young wolf - rayet ( wn - type ) star which experimented mass ejection episodes shortly before core collapse .
[ firstpage ] supernovae : general supernovae : individual ( sn 2005la , sn 2006jc ) |
in recent years much attention has been devoted to the study of relaxation processes of non - equillibrium charge carriers in strongly localized systems , where transport proceeds via phonon - assisted hopping , like photoexited charge carriers in band tails ( see e.g. @xcite-@xcite ) and anderson insulators ( see e.g.@xcite-@xcite ) . in such systems
particular small relaxation speeds are observed .
often the smallness of the relaxation speed is attributed to interaction effects .
however , also in strongly localized non - interacting electron systems long lasting relaxation processes are known to be the rule , and not the exception @xcite .
the theoretical investigation of such relaxation processes is notoriously difficult , since the system is strongly disordered and always in a transient state . in most problems
one is interested in time scales which are large as compared to the time a single hop needs . for such time
scales all quantities depend strongly on frequency , also for very low frequencies , so that the consideration of dispersive transport is vital . on the other hand in most problems of interest both spatial and energetical disorder exists . due to the latter fact the transitions are inelastic .
the inelastic character of the transitions , the relaxation , leads to a flow of energy from the electron system to the phonon system . due to disorder
this transport of energy is also dispersive .
therefore , dispersive energy transport and relaxation are intimately connected .
the investigation of the relaxation process requires both the consideration of dispersive particle transport _ and _ the consideration of dispersive energy transport .
the intricate physical situation manifests also in the equations , which have to be considered .
since the transport is inelastic already in the simplest approximation the investigation requires to find solutions to integral equations ( see e.g.@xcite- @xcite ) . due to the fact that the particle moves in an energy dependent density of states the kernel of these integral equations does not depend on the difference between the site energies only . the traditional method to handle the situation ,
the percolation theory , is not applicable here .
the effective - medium methods by movaghar and coworkers give , as pointed out by movaghar and coworkers , wrong results for systems at low temperature @xcite @xcite .
furthermore , beyond the markovian approximation the derivation of these integral equations represents itself an intricate problem . to our knowledge , beside the attempts by movaghar and coworkers , numerical investigations
( see e.g. , ) , and physical intuitive considerations ( see e.g. @xcite , ) , mainly markovian equations have been used ( see e.g. ) .
it is the aim of the present paper to fill this gap and to provide a formalism that can be used for studying relaxation phenoma of strongly localized , charge carriers far from equillibrium in the hopping regime taking into account both dispersive particle transport and dispersive energy transport . to this end
we focus , for the sake of definitness , on relaxation of photoxited , non - interacting , non - equillibrium charge carrires in band tails of , e.g. amorphous semiconductors , like amorphous si : h . on the first glance
, this problem seems to be rather special .
if the number of charge carriers exited is small fermi - correlation is negligible , so that one has to cope with linear rate equations .
the linearity of the transport equations is , of course , the basic ingredient in solving the problem
. a closer look on the problems of interest reveals , however , that most of the problems can be formulated in this way . clearly , how to achieve linearization depends on the experiment chosen , but , provided the number of charge carrires exited is small , always the smallness of the particle number can be invoked . in this case , the number of particles at a site is certainly not small as compared to its equillibrium value , but small as compared to unity .
after linearization the structure is , in principle , quite similar as for the case considered here . below we present the derivation of a framework for the consideration of relaxation phenomena due to phonon - assisted hopping at zero temperature .
we focus on the limit of strong localization , where dispersive effects are expected to be most pronounced . here
the disorder manifests in a strong dependence of the transport coefficients on frequency for very low frequencies .
consequently , in this regime the diffusion propagator can not be calculated from markovian transport equations , like those used e.g. in ref .. to simplify our integral equations we use the concept of quasi - elasticity , the particle changes its energy only by a small amount by one hop .
the method is applied to the study of dispersive energy transport .
we find that always two tendencies are present , if only the density of states not decreases with increasing energy .
first , there is a widening of the packet due to the statistical spreading , the dispersive energy transport .
second , there is narrowing of the packet due to decrease of the density of states with decreasing energy .
the overall evolution is determined by the interplay of these two tendencies .
we have studied both tendencies in two limiting cases , for a particle moving in a constant density of states and for a particle moving in an exponential density of states . in the first case
the impact of the density of states decrease is absent , so that only the statistical spreading is present . here
the packet evolves into a gaussian packet moving with constant drift velocity in energy space .
the packet width increases with time as @xmath0 .
the other result is obtained for exponential density of states .
here both tendencies are present .
we find , that the velocity of the packet is strongly slowing down with time . in that case
the mean square deviation of energy eventually becomes time - independent .
consequently , the motion of the packet in energy space is `` soliton - like '' .
the concrete results on the diffusion propagator , its time dependence and its moments for exponential densities of states are of relevance for photoluminiscense experiments on amorphous si : h .
we consider photoexited , localized charge carriers in band tails at zero temperture .
after exitation the charge carriers lower their energy by phonon - assisted hops between localized states .
since @xmath1 , only hops from higher to lower energy occur .
in this situation the charge carriers are in strong non - equilibrium .
we assume , that the number of excited charge carriers is small , and their energies are far from the fermi level , so that it is very unlikely that an electron jumps to a site already occupied .
consequently , we can neglect fermi - correlation . in this case
the electron transport can be described by the rate equation @xcite .
@xmath2 in calculating the transition probabilities we assume , that the electron - phonon coupling strength is weak , so that only one - phonon processes have to be taken into account .
then , at zero temperature , the transition probabilities are given by @xmath3 where @xmath4 here @xmath5 is the inverse of localized state radius , and @xmath6 is the phonon frequency .
the energy @xmath7 is the upper bound for the energy transfer by one hop .
note , that in the materials of interest @xmath7 can be much smaller as the debye - frequency , since not all phonons can interact with localized electrons equally well .
short wave - length phonons are ineffective since the electron - phonon coupling constant tends to zero for momenta @xmath8 with @xmath9 .
therefore , the effective upper phonon momentum is of the order @xmath10 , and not of the order of the inverse lattice constant of the host material .
furthermore , in disordered systems the high - energetic phonons are localized , and need not contribute to transport by necessity .
the first step function in front of the transition probabilities restricts the transitions to transitions between sites separated by at most @xmath11 in energy space .
thus , it decreases the energy relaxation speed . in impurity conduction , and in nearly all papers on relaxation of charge carriers in band tails , this step function is usually replaced by unity . in impurity conduction
this is quite reasonable , since it is assumed that hops are restricted within narrow stripe near the fermi level , which is small compared to debye energy . in the band - tail problem , however , we can see no reason for neglecting it in advance . to calculate the transport quantities of interest
, we have to calculate the diffusion propagator . in order to render the analytical calculations feasible
we introduce continuous coordinates .
the change of representation is defined by : @xmath12 where @xmath13 and @xmath14 . in this representation the rate equation ( [ b1 ] ) takes the form : @xmath15 where @xmath16 is determined by the equations : @xmath17 @xmath18 , \label{b7}\ ] ] @xmath19 the laplace - transformed equation is given by : @xmath20 where @xmath21 is the initial condition .
we will assume that @xmath22 is a function @xmath23 , so that @xmath24 .
equation ( [ b9 ] ) can be solved using the green s function method .
the solution is given by : @xmath25 the green s function satisfies the equation : @xmath26 note , that , due to probability conservation , the green s function @xmath27 and the probability @xmath28 satisfy the relations : @xmath29 @xmath30
in order to calculate the configuration average , we assume , that the sites are distributed homogeneously in space .
the distribution of site energies is supposed to be given by some distribution function @xmath31 .
accordingly , the average of any quantity , depending on the energy and positions of sites , is given by : @xmath32 where @xmath33 is the volume of the system .
the application of the averaging procedure to the structural factor @xmath34 serves , in particular , as a definition for the density of states , i.e. @xmath35 products of the structural factor @xmath34 are averaged according to : @xmath36 using these definitions , the configuration average of eq.([b10 ] ) can be calculated diagrammatically @xcite .
the diagrammatic method leads to the following set of equations for the calculation of the configuration average @xmath37 of the electron density @xmath38 @xcite , ( see figs [ fig1][fig4 ] for illustration ) : @xmath39 @xmath40 @xmath41 @xmath42 @xmath43 \ , .
\label{c8}\ ] ] here @xmath44 .
a detailed derivation of the set of integral equations is given in appendix [ ap1 ] .
given the system of integral equations ( [ c5])-([c8 ] ) , the main problem is to find an approximate self - consistent solution to it .
the situation is quite similar to that of the calculation of the equilibrium conductivity in a disordered system @xcite .
there an approximate solution of this system could be found by introducing a proper decomposition of the function @xmath45 , the diffusion propagator , into short- and long - wavelength limit , according to : @xmath46 where @xmath47 was a frequency dependent parameter that could be related to the critical hopping length @xmath48 via the equation : @xmath49 and @xmath50 was an energy dependent function determined by the principle of detailed balance .
this decompostion originated from the notion , that the integrals in the integral equations are governed by products of transition probabilities and diffusion propagators , and , for strongly localized electrons , the latter quantities are short ranged functions as compared to the transition probabilities . using the decomposition ( [ em1 ] ) , the effective medium approximation reduces to the repacement of @xmath45 by @xmath51 in the calculation of the effective transition probability @xmath52 ( eq.([c7 ] ) ) and of the irreducible block @xmath53 ( eq.([c8 ] ) ) . here in the band - tail problem
we use the same philosophy .
we first decompose the diffusion propagator in two parts , according to : @xmath54 then , to investigate the consequences of this decomposition , we insert eq.([em3 ] ) into eq.([c7 ] ) . performing a partial summation
we obtain : @xmath55 where the renormalized transition probabilities @xmath56 are given by eq.([b7 ] ) , with @xmath57 replaced by : @xmath58 note that the renormalized transition probability depends now on @xmath59 via the function @xmath60 . at this point
the introduction of the function @xmath60 , the effective medium , still seems to be rather arbitrary .
however , if we now choose the effective medium in such a way , that the integrals over @xmath61 vanish , the renormalized transition probabilities yield the exact solution of the diffusion problem . in that case
the function @xmath53 turns into : @xmath62 so that eq.([c4 ] ) can be cast into the form : @xmath63 therefore the function @xmath64 can be identified with the diffusion propagator .
note , that although the general philosophy is quite the same as in equilibrium , the situation is much more intricate here , the principle of detailed balance is absent , since we are dealing with the situation at zero temperature , and so it ca nt determine the energy dependence of the effective medium .
moreover , we expect the critical hopping length to depend somehow on the position of the electron in the tail .
thus , in contrast to equilibrium , @xmath65 is not only a frequency dependent parameter but also an energy dependent function , which has to be determined self - consistently .
for the moment we put aside the question of the determination of the effective medium to elaborate further on the consequences of the renormalization of the transition probabilities . to this end
we focus on the diffusion propagator . the equation for the diffusion propagator is given by eq.([c1 ] ) .
in effective medium approximation , when calculating the irreducible part @xmath52 , only the lowest order contribution to @xmath52 with respect to @xmath66 is taken into account , so that @xmath67 . in that approximation
the equation for the diffusion propagator in momentum representation reads : ' '' '' ' '' ''
@xmath68 of course , as we do nt know how the effective medium looks like so far , and moreover , as the equation is a complicated integral equation , we ca nt find a solution .
the fact , that the effective medium is a function of energy , makes the problem much more complicated .
progress can only by achieved if we can find arguments to simplify the equation considerably . to this end
we focus on the renormalized transition probability .
according to the eqs .
( [ b2 ] ) and ( [ em5 ] ) , it is given by : @xmath69 owing to the step functions in front of the transition probability , the energy integrations are restricted to intervals of length @xmath11 . taking this fact into account the integral equation for the calculation of the diffusion propagator takes the form : @xmath70 further we assume that the diffusion propagator , the transition probability @xmath71 , and the density of states @xmath72 are slowly varying functions on intervals of length @xmath73 , so that the integrand can be expanded with respect to @xmath74 . doing
so we obtain : @xmath75 f\left ( { \bf q}|\varepsilon ^{\prime } , \varepsilon \right ) + \frac
12% \omega ^2\frac \partial { \partial \varepsilon } \left [ f\left ( { \bf q}% \label{p4}\ ] ] ' '' '' ' '' '' 2 we terminate the expansion after the first derivative with respect to energy .
this term describes the biased motion of the particle from sites of higher to sites of lower energy . for finite temperatures one would have to replace @xmath76 in the last term with @xmath77 which would describe the energy diffusion current . at zero temperature
there is no input of energy from the phonon - system into the particle .
therefore , we expect the additional term to be rather small , and can be neglected . to investigate the transport properties ,
we restrict our attention with the long - wavelength limit . in that case
the elastic contribution yields the diffusion coefficient for particle diffusion .
the term , containing the derivative with respect to energy is finite for @xmath78 . if we are interested only in the long - wavelength limit , we can put here @xmath79 , since the remaining terms are of higher order with respect to @xmath8 and @xmath80. since a nonzero momentum in this term couples particle diffusion to energy transport , this approximation corresponds to a decoupling of these two processes .
then , in the long - wavelength limit , we obtain : @xmath81 \ , , \label{p5}\end{aligned}\ ] ] where : @xmath82 and : @xmath83 eq.([p5 ] ) is easily solved .
it s solution is : @xmath84 \ , .
\label{p8}\ ] ] explicit expressions for the transport coefficients can be found in appendix [ ap2 ] .
now , at this stage , the validity of our quasi - elastic approximation requires , that the second derivative terms are small as compared to terms with first derivatives with respect to energy .
this requirement imposes the following restrictions on the transport coefficients @xmath85 and @xmath86 , the frequency @xmath59 and the momentum @xmath8 : @xmath87 @xmath88 @xmath89 the applicability of the quasi - elastic approximation was also discussed in ref.@xcite .
there it was concluded that this approximation should be inapplicable . to substantiate this statement
numerical calculations were invoked . however , in interpreting these data it has to be taken into account , that they have been obtained using a model , neglecting a weighting of the transition probabilities according to the number of phonons emitted , so that hops between nearly isoenergetic sites were treated as likely as hops from the very top to the very bottom of the tail .
consequently the discussion in ref.@xcite applies only to system with sufficient strong electron - phonon interaction , but not to systems with weak electron - phonon interaction . for extraordinary deep hops to contribute to the diffusion propagator
they should be characteristic for an ensemble of electrons .
this is , however , not expected and in the experimental data , e.g. on amorphous si : h @xcite , not observed . for these reasons the discussion in ref.@xcite does not apply to our model .
before establishing a self - consistency equation , we first elaborate further on the physical content of our diffusion equation . to this end
let us first have a closer look on the coefficients @xmath90 and @xmath91 .
imagine , that we have a particle initially located at ( @xmath92 ) .
then the initial condition is @xmath93 . according to our approximation ,
the motion of the charge carrier is composed of two contributions : particle diffusion between isoenergetic sites , and relaxation in energy space .
characteristics of these two processes are the mean square displacement and the energy relaxation speed .
it turns out , that both can be calculated from the function : @xmath94 \nonumber \\
= \frac{\theta \left ( \varepsilon _ 0-\varepsilon \right ) } { v\left ( \varepsilon , s\right ) } \exp \left [ -s\int\limits_\varepsilon ^{\varepsilon _ 0}\frac{d\varepsilon _ 1}{v\left ( \varepsilon _ 1,s\right ) } \right ] \ , , \label{pi0}\end{aligned}\ ] ] which is just @xmath95 .
we define the mean energy of the particle by : @xmath96 using our diffusion propagator , this equation can be rewritten in the form : @xmath97 and from the diffusion equation we obtain after integration by parts : @xmath98 therefore , in general , the time dependence of the mean energy is given by complicated integrals .
these integrals simplify considerably in two limiting cases : in the absence of dispersive energy transport , and for energy independent @xmath99 . in the absence of dispersive energy transport ,
i.e. in the markovian limit in which the transport coefficients are independent of @xmath59 , the integrals can readly be calculated in time representation . in the latter situation
we simply obtain : @xmath100 where @xmath101 is defined by : @xmath102 therefore we obtain : @xmath103 consequently , @xmath85 is the velocity of energy relaxation , and @xmath104 is the instantenous position of the particle in energy space . in general , however , in disordered systems the coefficient @xmath99 depends on @xmath59 , so that _ energy transport is dispersive_. in that case the integral in eq.([pi11 ] ) can only be calculated easily if @xmath85 is independent of energy .
then , owing to probability conservation , @xmath105 is obtained .
the mean square displacement , defined by : @xmath106 can be obtained from similar arguments . in fourier representation
this equation can be written in the form : @xmath107 the derivative of the diffusion propagator can be calculated using the diffusion equation . doing so , we obtain : @xmath108 again eq.([pi5 ] ) simplifies only in two special cases , in the markovian limit and for energy independent diffusion coefficients .
whereas in the first case @xmath109 is obtained , we have @xmath110 in the latter situation . below
we shall see , that energy independent transport coefficients are obtained for constant density of states only .
so far we have only investigated the consequences of the renormalization . in order to complete the approximation scheme , we still have to calculate the effective medium itself . of course
, we can not calculate the effective medium exactly , since this would amount to find an exact solution to the diffusion problem .
rather we shall try to calculate @xmath60 self - consistently .
the transport coefficients are properties of the green s function @xmath45 , the diffusion propagator .
thus , in order to find an equation for @xmath111 we should relate the transport coefficients to @xmath52 , the irreducible part of the diffusion propagator .
the diffusion coefficient comprises only elastic contributions , however , in the relaxation problem a proper description of inelastic processes is vital .
moreover , as shown in the previuos section , the characteristics for particle and energy transport can already be calculated from the function @xmath112 , the @xmath78 limit of the diffusion propagator .
therefore , in establishing a self - consistency equation we should focus on those characteristics that are important in that limit , that is the energy relaxation speed @xmath85 .
the equation for the diffusion propagator for @xmath79 is given by : @xmath113 if we compare eq.([sc1 ] ) with eq.([p5 ] ) we deduce that @xmath114 now we decompose @xmath85 , as defined in eq.([sc2 ] ) , into two parts , one part that contains only the effective medium approximation of @xmath52 : @xmath115 and a part @xmath116 , that contains the deviations @xmath117 .
self - consistency requires : @xmath118 @xmath119 , as defined by eq.([sc3 ] ) , is in accordance with the definition ( [ p7 ] ) , taking into account eq.([t3 ] ) and the inequality @xmath120 ( eq . ( [ p11 ] ) ) ,
owing to which contributions proportional to this parameter are negligible .
we now focus on the self - consistency equation ( [ sc4 ] ) . while eq.([sc3 ] ) contains only the effective medium contribution to the diffusion propagator , @xmath121 is a functional of the effective medium @xmath65 and the deviation @xmath66 . by construction ,
it is at least linear in @xmath66 .
if this equation could be solved exactly , an exact solution to the diffusion problem could be found within quasi - elastic accuracy . in practise
this is not possible , and , therefore , we depend on further approximations . to simplify this equation ,
we take into account only the lowest order contributions to this equation with respect to @xmath66 , i.e. we linearize @xmath121 with respect to @xmath66 and require that the first order contribution vanishes .
this approach is quite close to the usual cpa - philosophy , in which vanishing of the t - matrix is required in its lowest order approximation . using this procedure
we obtain the following self - consistency equation : @xmath122 where @xmath123 and @xmath124 are simply numbers .
a detailed derivation of this equation is given in appendix [ ap3 ] .
in deriving the self - consistency equations we have imposed further restrictions on the effective transition probabilities , which determine the range of its applicability .
these restrictions can be formulated most conviniently using the dimensionless critical hopping length @xmath125 , related to @xmath111 by ( see appendix [ ap2 ] ) : @xmath126 in terms of @xmath127 , the inequalities ( [ p11 ] ) and ( [ p12 ] ) , used in the derivation , read ( prime is derivative with respect to @xmath128 ) : @xmath129 @xmath130 in addition , when calculating the integrals , @xmath131 was used . a closed solution to the self - consistency equation
can only be found in the limit @xmath132 .
there we obtain : @xmath133 ^{1/d}}\left [ \frac{2da}{s\left ( d\right ) } \right ] ^{1/d } , \label{sc6}\ ] ] where @xmath134 is the solid angle in @xmath135 dimensions . for @xmath59 satisfying ( [ sc5c ] ) , eq.([sc5 ] )
can be cast into the form : @xmath136 \exp \left [ \rho _ c\left ( \varepsilon , 0\right ) -\rho _ c\left ( \varepsilon , s\right ) \right ] = \frac s{\omega \left ( \varepsilon \right ) } \ , , \label{sc8}\ ] ] where : @xmath137 according to eq.([sc8 ] ) , the critical hopping length decreases with increasing frequency .
note , that the structure of the equation ( [ sc8 ] ) for the calculation of the dispersion of the critical hopping length , obtained here , agrees completely with that obtained for the critical hopping length in calculating the equilibrium conductivity @xcite .
the dispersion of the transport coeffcicients is determined completely by the dispersion of the critical hopping length . for small @xmath59 the frequency dependence
preexponetial factors can be ignored , so that from eq.([sc8 ] ) explicit equations for the transport coefficients can be obtained .
they are given by : @xmath138 @xmath139 the formal solution of these equations is given by the lambert s w - function @xmath140 , defined by the equation @xmath141 .
using the lambert s function we can write : @xmath142 @xmath143
a constant density of states , although of not much physical relevance , gives us the unique opportiunity to study pure dispersive energy transport . here
both @xmath144 and @xmath145 are independent of energy .
consequently , eq.([pi0 ] ) can readly be integrated .
the integration yields : @xmath146 \ , .
\label{cds1}\ ] ] the time - dependence of this function can be obtained by the inverse laplace - transformation . using eq.([sc11 ] ) to change the integration variable from @xmath59 to @xmath147 , the inverse laplace - transform of eq.([cds1 ] )
may be written as : ' '' '' ' '' '' @xmath148 \ , , \label{cds2}\ ] ] with the properly chosen integration contour @xmath149 . at large enough @xmath150 , that is for @xmath151 , this expression , using the saddle - point method , gives simply the gaussian packet : @xmath152 \ , .
\label{cds3}\ ] ] ' '' '' ' '' '' 2 now let us consider the time dependence of the energy relaxation . according to eq.([pi15 ] )
the velocity of energy relaxation is : @xmath153 again , the time dependence of the energy relaxation speed can be calculated using asymptotics .
details of the calculations are presented in appendix [ ap5 ] , where it is shown that : @xmath154 ^{-2}\,,\;{\rm as}\;\omega t\ll 1\ , .
\end{array } \right .
\label{cds4}\ ] ] note , that the problem of the energy relaxation in the case of constant density of states ( @xmath99 is independent of @xmath128 ) is completely equivalent to one of the nonmarkovian charge transport in strong electric field @xmath155 , when the diffusion ( described by the second coordinate derivate ) is totally neglected .
one has only to replace @xmath156 , @xmath157 being the mobility@xcite . in the context of dispersive particle transport ,
the regimes @xmath158 and @xmath151 are the regimes of anomalous and normal `` diffusion '' , respectively .
the main difference between both is in that while the sites are ususally distributed homogeneously in space , the density of states is usually an increasing function of energy .
the calculation of the time dependence of the energy distribution function for an arbitrary density of states on the basis of the eqs .
( [ p5])-([p7 ] ) , ( [ sc10 ] ) and ( [ sc11 ] ) turns out to be a quite intricate problem . a tool that can be utilized in tackeling the problem is the saddle point approximation . how to apply the saddle point approximation for the calculation of the quantities of interest for an arbitrary density of states
is shown in appendix [ spa ] .
below we focus on the exponential density of states , which is relevant , e.g. , for amorphous si : h .
we assume that the density of states is given by : @xmath159 to simplify the notations we use the abbreviations : @xmath160 @xmath161 then , the eqs.([t3 ] ) , ( [ sc6 ] ) and ( [ sc9 ] ) can be cast into the form : @xmath162 \ , , \label{aes3}\ ] ] @xmath163 \ , , \label{aes4}\ ] ] @xmath164 where : @xmath165 ^{1/d}}\left [ \frac{2da}{s\left ( d\right ) } \right ] ^{1/d}\ , .
\label{aes6}\ ] ] using these equations it follows from the formulas derived in appendix [ spa ] that the time dependence of the mean squared deviation and the time dependence of the mean energy are given by @xmath166 \ , , \label{aes7}\ ] ] @xmath167 \ , , \label{aes8}\ ] ] where @xmath168 note , that , according to the saddle - point approximation , the distribution is gaussian .
furthermore , the dispersion is constant for @xmath169 .
consequently , for time scales in line with the applicability of the saddle - point approximation , the motion of the energy packet is `` soliton - like '' , that is the packet moves without distortion .
the applicability condition for the saddle - point ( [ sp8 ] ) requires : @xmath170 because of @xmath171 grows with @xmath150 , the saddle - point method , and , consequently , all the results of this chapter , becomes invalid , at least at sufficiently large @xmath150 .
the physical meaning of the condition ( [ aes10 ] ) is the following : at @xmath169 the width of the electron s energies distribution is @xmath172 .
because of @xmath173 , the variation of @xmath174 is : @xmath175 , and , because @xmath176 , we have : @xmath177 .
thus , the condition ( [ aes10 ] ) is just the one of the variation of electron s `` velocity '' across the distribution is smaller , then the velocity itself .
even if the initial condition are such , that the condition ( [ aes10 ] ) is fulfilled , and a gaussian distribution is formed , at some moment of time @xmath178 becomes of the order of @xmath179 , which definitely should results in some deviation of the distribution function from its symmetric , gaussian form .
for the exponential density of states , distribution function in the laplace representation , eq.([pi0 ] ) , using eqs.([sc13],[aes3]-[aes5 ] ) , may be written as : ' '' '' ' '' '' @xmath180 \ , = \frac{{\cal w}\left(s\rho e^\rho \right ) } { { \bar\omega } \rho s } \exp \left [ -\frac 1{\bar\omega } \int\limits_{\rho _
0}^\rho dx\frac{{\cal w}\left ( sxe^x\right ) } { x^2}\right ] \ , , \label{dlt1}\ ] ] where it was set @xmath181 and @xmath182 by appropriate choice of time and energy units . under the same conditions , which were used in the derivation of the above formula , the following approximation is valid ( see appendix [ ap6 ] ) : @xmath183 + o\left ( \frac{{\cal w}}{% \rho ^3},\frac{{\cal w}^3}{\rho ^3}\right ) \ , .
\label{dlt2}\ ] ] when calculating distribution function at sufficiently large times , the characteristic values of @xmath59 , giving the main contribution into inverse laplace integral become so small , that one can suppose @xmath184 , and : @xmath185 . under this condition
the distribution function approach initial conditions independent shape : @xmath186 , @xmath187 , @xmath188 \right\ } \ , .
\label{dlt3}\ ] ] let us consider , first , momenta of the distribution at the large times . in laplace - representation
the momenta are defined by the equation : @xmath189 ' '' '' ' '' ''
2 we shall omit further on irrelevant time shift multiple @xmath190 .
it turns out , that for @xmath191 , the integrand of eq.([dlt4 ] ) has a maximum , located at : @xmath192 however , if one would try to use the saddle - point method in evaluation of integrals like ( [ dlt1 ] ) , the results would be incorrect at @xmath193 , since the saddle point criterion is not fulfilled .
however , at @xmath194 , momenta can be calculated as an expansion on the powers of small paramenter @xmath195 . details of the calculations are presented in appendix [ ap6 ] . as a result , we have for the first momentum , or the mean energy : @xmath196 where @xmath197 , and @xmath198 is eulers s constant . for the distribution
s dispersion @xmath199 we obtain : @xmath200 and for the third central momentum @xmath201 we have : @xmath202 \frac{\bar\omega } { \ln \left ( \sqrt{\bar\omega } t\right ) } \ , , \label{dlt7}\ ] ] figures [ fig5 ] , [ fig6 ] , and [ fig7 ] show mean energy , @xmath203 , mean square deviation of energy @xmath204 , and dimensionless coefficient of asymmetry : @xmath205 respectively , plotted versus @xmath206 .
there values , obtained numerically , using the distribution function ( [ dlt3 ] ) , are compared with their corresponding asymptotic forms , following from eqs.([dlt5])-([dlt7 ] ) . moving along the same line
, the whole distribution function may be reconstructed at large times , if to calculate all its momenta .
however , it may be obtained in more simple way .
namely , when @xmath207 , one may replace @xmath59 within arguments of @xmath208-functions in eq.([dlt3 ] ) with some @xmath209 , @xmath210 ( a kind of argumentation may be found in appendix [ ap7 ] ) .
then the integral can easily be calculated to yield : @xmath211 \right\ } \ , .
\label{dlt9}\end{aligned}\ ] ] ' '' '' ' '' '' ' '' '' ' '' '' 2 one can easily show ( see appendix [ ap7 ] ) , that asymptotic expressions for momenta ( [ dlt5]-[dlt7 ] ) may be entirely reproduced from eq.([dlt9 ] ) , if to set @xmath212 ( @xmath198 is the euler s constant ) .
figures [ fig8]-[fig10 ] show some plots of the distribution function ( [ dlt3 ] ) , compared with its asymptotic form ( [ dlt9 ] ) .
this latter can be further symplified , if to introduce @xmath213 : @xmath214 as we shall see , the body of the distribution at a given time corresponds to values of @xmath215 , @xmath216 .
then , one can use approximation : @xmath217 . then
, we have in the exponent @xmath218 , which can be replaced by @xmath219 , where @xmath220 . as a result , at large time and small @xmath221
the distribution function turns into : @xmath222 \,,\;\varepsilon < \varepsilon _ t \\
1\,,\;\varepsilon > \varepsilon _
t \end{array } \right .
t=\ln \ln t\ , . \label{dlt11}\ ] ] at @xmath223 it was supposed @xmath224 .
to understand , what happens at the intitial stage of the distribution s evolution , let us consider the behaviour of @xmath225 at @xmath226 .
one can write from eqs.([sp0],[sc13],[aes3]-[aes5 ] ) : @xmath227 where the new integration variable @xmath228 was introduced instead of @xmath59 , and the integration by parts was performed .
this integral may be evaluated using the saddle - point method .
the saddle - point equation , @xmath229 , gives us saddle point value @xmath230 , that is @xmath231 at @xmath232 , and @xmath233 as @xmath234 . to ensure the correctness of the saddle point approximation
, one should require the parameter@xmath235 to be small .
while this is true at @xmath234 , this parameter appears to be @xmath236 at @xmath237 .
therefore , the result for @xmath237 is correct up to the multiple of the order of 1 only .
we have : @xmath238 in a similar fasion one can calculate the energy derivative at @xmath239 : @xmath240 \exp \left ( \tau ze^z\right ) \
, , \label{ise3}\end{aligned}\ ] ] where the term @xmath241 within the square brackets was neglected due to assumption @xmath242 . at @xmath234
this integral may be readily evaluated in the saddle point approximation , which yields : @xmath243 when @xmath237 , one have to perform integration by parts first , which gives : @xmath244 \exp \left ( \tau ze^z - z\right ) \,.\end{aligned}\ ] ] then one have in the saddle point approximation : @xmath245 at sufficiently small @xmath150 the above expression is positive , which means that the distribution is `` sticked '' near @xmath246 , that is , it monotoneously decreases as @xmath247 increases . at some @xmath248 , @xmath249 changes its sign .
if @xmath250 , this corresponds to @xmath251 , otherwise @xmath252 .
this correspons just to the moment of time , when this distribution separates from the intial point , its maximum is at @xmath253 . for @xmath254 the width of the distribution on the energy scale may be estimated as : @xmath255 this formula is valid if @xmath256 , up to the times corresponding either @xmath257 if @xmath258 , or @xmath259 otherwise . in the former case
the distribution width becomes of the order of @xmath260 just before it separates from the initial point .
after this , the distribution , as it was shown in previous subsection , widens till @xmath261 , and the gaussian packet moves downward , untill its center reaches the value , corresponding to @xmath262 . in the latter case
the packet s width is @xmath263 at the very moment of its `` unsticking '' . in both cases , to consider subsequent evolution , one has to use some other approximation instead of saddle point one .
in the paper we have presented an effective method which permitts the investigation of relaxation phenomena of localized charge carriers far from equillibrium due to phonon - assisted hopping at zero temperature . from the point of view of the formalism
the main equations are the equations for the calculation of the diffusion propagator ( [ p5 ] ) , ( [ p6 ] ) and ( [ p7 ] ) , and the equations ( [ sc9 ] ) , ( [ sc10 ] ) and ( [ sc11 ] ) , which determine the dispersion of the transport coefficients .
these equations show , that both particle transport and energy transport are dispersive . the equations ( [ sc10 ] ) and ( [ sc11 ] ) lead to a strong dependence of the diffusion constant @xmath264 and the velocity of energy relaxation @xmath265 on frequency @xmath59 already for low frequencies . to our knowledge
, these equations have not been dervied in the literature so far , for systems far from equillibrium .
in fact , in the literature mainly frequency independent transport coefficients can be found ( see e.g. @xcite and references therein ) .
the strong dependence of the transport coefficients on @xmath59 results in a non - markovian equation for the calculation of the diffusion propagator ( [ p5 ] ) .
this discriminates our equations from the markovian integral equations used in the refs . @xcite-@xcite .
the latter equations can , in principle , be obtained from our leading equation ( [ b11 ] ) by neglecting statistical correlation , so as to average every factor independently .
using our effective - medium method , we have investigated the relaxation of charge carriers in band tails at @xmath1 . according to our results ,
energy relaxation is connected with dispersional transport .
even if we have no real diffusion in the system , since the temperature is zero , we have some spreading of the energy distribution with time . for a constant density of states
the situation is completely equivalent to the one in the problem of the electron s motion in a disordered system , subjected to an electric field @xcite .
the main difference between the above mentioned problem and the energy relaxation is in that , while in the former case the sites are distributed homogeneously in space , in the energy relaxation problem the density of states is a decaying function of energy for most physical systems . in this situation
we arrive at the picture that the particles of a package with lower energies move slower than particles with higher energy .
this leads to the opposite tendency : at first , there is a slowing of the package spreading .
the time dependence of the dispersion , being linear at the first stage of the evolution , is later slowing down . later on ,
different possibilities exist , dependend on the particular type of the energy dependence of the density of states .
we performed a detailed investigation for the exponential density of states for two time regimes .
if the variation of the velocity @xmath85 in the energy space across the distribution is smaller than the velocity itself , that is if @xmath266 , we arrive at the situation of a package , moving steadily down along the energy axis without any deformation .
dispersion becomes time independent ( see eq.([aes7 ] ) ) .
however , since the particles are sinking down this condition becomes violated when time goes by .
the parameter @xmath267 , being small at the first stage of the evolution , is getting large .
when this parameter is larger than one , the steady motion condition becomes violated again .
the package , being of gaussian form before , undergoes some restructuring to another , non - gaussian stable form , with its width lower than before by some numeric factor of the order of 1 ( eq.([dlt6 ] ) ) . in our pictures , figs [ fig7]-[fig10 ] , it is clearly seen , that the package becomes non - gaussian .
this result remains valid until the very moment the quasi - elasticity conditions break down ( @xmath268 ) . for the exponential density of states
we have found the packet to move as @xmath269 , i.e. , its motion is strongly slowing down with time , the packet becomes , roughly speaking , almost stopped .
this type of behaviour may be called `` glassy '' , because of the overall time scale for the packet evolution ( governed by the exponential function of the large parameter @xmath270 ) , becomes huge .
for example , when @xmath271 , this time scale approaches @xmath272 and more ( see fig .
[ fig10 ] ) ! the main simplification used in our paper is the quasi - elastic approximation .
this approximation relies on the smallness of the upper bound of the energy transfer to the phonon system by one hop . for localized electrons
, this upper bound can be much smaller as the debye energy of the host material , since not all phonons can interact with localized electrons equally well . for localized electrons the electron phonon coupling constant approaches zero , for phonons with wave - vector @xmath273 . thus high energetic phonons are less effective
. only phonons with energies @xmath274 , where @xmath123 is the lattice constant of the host material , are effective .
furthermore , in disordered systems the high energetic acustical phonons are localized , and thus need not contribute to transport by necessity . nevertheless , the question whether the quasi - elastic approximation is applicable depends much on the material of interest .
if , however , we compare our result for the time dependence of the mean energy ( [ aes8 ] ) in an exponential density of states with those existing in the literature @xcite , we also find agreement for @xmath275 , which indicates that our results are , at least qualitatively , of wider validity .
unfortunately , to our knowledge , in the literature there are no further results on the width of the energy distribution available to compare with .
from our point of view the main question remained open is how the results , obtained for the exponential density of states , may be generalized for other types of energy dependencies .
one can imagine , e.g. , that for densities of states decaying with decreasing energy slower than exponential , the dispersion is growing , but under some sublinear law .
also , in the opposite case , in a density of states decaying faster than exponential , we would possibly have a dispersion tending to zero for large times .
it should , however , also be mentioned that it is not completely clear under which situation zero temperature results can be applied to systems at finite temperature . in our approach
we assumed that @xmath1 .
however , in a real system , when the carriers are sinking down , the criterion for the temperature to be treated as zero , is violated at every finite value of @xmath276 at some moment of time , since the contribution of hops to sites with higher energy values becomes more and more comparable with one of downward hops .
therefore , the consideration of temperature becomes vital .
work in this direction is in progress . on of the authors ( o.b . ) would like to thank prof .
p. thomas for drawing his attention to the photoluminscense problem and stimulating discussions . ' '' '' ' '' ''
here we derive the set of integral equations ( [ c4])-([c8 ] ) .
the same method has already been used in ref.@xcite . according to eq.([b10 ] ) , the configuration average of the electron density @xmath277 is given by : @xmath278 where @xmath279 is the diffusion function . according to eq.([b11 ] ) , the diffusion function is given by : @xmath280 diagrammatically this series can be represented as depicted in fig.[fig1 ] . in this picture every full dot represents a potential @xmath16 , which depends on disorder via the structur factor @xmath34 .
the expansion starts with a single structural factor @xmath34 , depicted by the empty dot .
note that all disorder is comprised into the structure factor @xmath34 . in order to calculate the configuration average
, we average the series ( [ ca3 ] ) term by term .
the resulting expansion can be depicted in the usual form , as in impurity scattering problem ( see fig.[fig2 ] ) .
one can see from the picture , that the set of all diagrams can be decomposed into two subsets , the set @xmath53 containing all diagrams connected with the empty point and the set @xmath45 containing all other diagrams .
owing to this decomposition configuration averaged diffusion function can be written as : @xmath281 from which eq.([c4 ] ) follows .
note that @xmath45 is nothing but @xmath282 .
again the class of diagrams contributing to @xmath45 can be decomposed into reducible and irreducible diagrams .
all irreducible diagrams can be comprised into a function @xmath52 .
doing so , we obtain eq.([c5 ] ) .
note , that , although the introduction of the irreducible part @xmath52 is parallel to the introduction of the self - energy it has to be stressed , that its physical interpretation is completely different . in the present context
@xmath283 is the true transition probability from @xmath284 to @xmath174 .
it is @xmath285 that has to be calculated and compared with the experimental situation and not the bar transition probability , as suggested in many papers ( see e.g. @xcite ) .
diagrams , contributing to @xmath52 are depicted in fig.[fig3 ] .
the set of diagrams depicted in fig .
[ fig3 ] can be generated by means of the propagator @xmath286 , defined by the equation ( [ c7 ] ) , using eq.([c6 ] ) .
the propagator @xmath286 can also be used to generate the irreducible block @xmath53 .
diagrams contributing to @xmath53 are depicted in fig.[fig4 ] . from the picture it follows , that @xmath53 is given by eq.([c8 ] ) .
according to eq.([p7 ] ) , @xmath99 is given by : @xmath287 \int d{\bf r}\frac 1{1+\exp \left [ 2\alpha r-\rho _ c\left ( \varepsilon , s\right ) \right ] } \ , , \label{t1}\ ] ] where : @xmath288 we assume , that @xmath127 is large , so the integrals can be calculated in the limit @xmath289 .
this assumption is justified a posteriory by explicit calculation of @xmath127 . in limit
@xmath289 , the integrand reduces to a step function , so that the integrals can be calculated easily .
doing so , we find : @xmath290 ^d\nu \exp \left [ -\rho _ c\left ( \varepsilon , s\right ) \right ] \ , , \label{t3}\ ] ] where @xmath134 is the solid angle and @xmath135 is the spatial dimension . from the calculation and the result
it is to be seen , that @xmath125 is to be identified with the dimensionless critical hopping length at energy @xmath128 .
the same procedure is applied to the calculation of @xmath90 , defined by eq.([p6 ] ) . using the same approximations ,
we obtain : @xmath291 ^{d+2}\nu \exp \left [ -\rho _ c(\varepsilon , s)\right ] \ , .
\label{t4}\ ] ] thus @xmath90 and @xmath99 differ only in preexponential factors .
we anticipate that for @xmath132 we obtain : @xmath292 ^{1/d}}\left [ \frac{2da}{s\left ( d\right ) } \right ] ^{1/d}\ , , \label{t5}\ ] ] where @xmath123 is a number and @xmath134 is the solid angle
. a different form for the diffusion coefficient can be found in ref.@xcite .
the expression given in this paper differs from our expression essentially only in that @xmath11 is replaced by the tailing parameter . in looking on this difference one should , however , take into account , that the model used in that paper is different .
furthermore , in ref.@xcite the transport coefficient are independent of @xmath59 .
it is therefore not completely clear to which range of time they apply .
clearly , they can not determine the evolution for all times .
here we derive the self - consistency equation ( [ sc5 ] ) . in order to work out
explicitely the first order contribution to the self - consistency equation ( [ sc4 ] ) we first need the first order correction to @xmath52 .
it is given by : @xmath293 if we insert the expression for @xmath294 , we obtain : @xmath295 \tilde{w}\left ( \rho _ 1,\rho ; s\right ) -{\cal n}\left ( \varepsilon _ 1\right ) \tilde{w}% \left ( \rho ^{\prime } , \rho _
1;s\right ) \left [ \tilde{f}\left ( \rho _ 1,\rho \right ) -\tilde{f}\left ( \rho ^{\prime } , \rho \right ) \right ] \tilde{w}% \left ( \rho , \rho _ 1;s\right ) \right\ } \ , . \label{as2}\end{aligned}\ ] ] thus , at this stage the self - consistency equation takes the form : @xmath296 in order to eliminate @xmath66 , we have to replace @xmath297 , where @xmath45 is the effective medium approximation of the diffusion propagator .
let us first work out the local contribution .
replacing @xmath117 by @xmath298 , the first and the third term of eq.([as3 ] ) cancel each other .
the 4th term is zero , taken into account that the delta function of the effective medium is multiplied by its argument .
thus , when @xmath66 is replaced by @xmath299 , only the 2nd term of eq.([as3 ] ) survives . to simplify this term we take into account that @xmath300 where @xmath301 is a dimensionless function .
consequently , we obtain @xmath302 ^2 \approx -\int d{\bf q}\left [ \tilde{w}\left ( { \bf q}|\varepsilon , s\right ) \right ] ^2\frac{\omega ^2}2f\left ( \varepsilon , s\right ) { \cal n}\left ( \varepsilon \right ) \nonumber \\ & = & -\frac{1}{2}\omega^2\left(\frac{2\alpha}{\rho_c(\epsilon , s)}\right)^d { \cal n}(\epsilon)f(\epsilon , s){\tilde w}^2(0|\epsilon , s)\int d^dx\phi^2(x ) \ , . \label{as4}\end{aligned}\ ] ] here terms proportional to @xmath303 have been neglected .
let us now focus on the contribution of the regular part of the diffusion propagator to the self - consistency equation .
owing to the step functions in @xmath304 an @xmath305 , the first and third term in eq.([as3 ] ) are zero , when @xmath66 is replaced by @xmath45 .
thus we are left with : @xmath306 the range of integration in ( [ as5 ] ) is determined by the step functions in @xmath45 and @xmath307 .
owing to these step functions , the energy integrations extend at most over intervals of length @xmath73 , so that the quasi - elastic approximation can again be applied to the effective - transition probabilities and the density of states .
the diffusion propagator enetering eq.([as5 ] ) , however , can not be dealt with in this way since for arbitrary @xmath8 the derivatives of the diffusion propagator are not small as compared to the diffusion propagator itself . therefore , another procedure is needed . to simplify this expression further we consider eq.([p1 ] ) .
if terms small with respect to @xmath308 and @xmath309 are neglected the function @xmath310 can be introduced , that satisfies the equation @xmath311.\end{aligned}\ ] ] then the expression ( [ as5 ] ) can be cast into the form @xmath312\ ] ] for @xmath132 eq.([as5b ] ) does not contain any physical parameter .
it only leads to the determination of the function @xmath313 which satisfies the equation @xmath314.\end{aligned}\ ] ] provided , we restrict our consideration to small frequencies in deriving the self - consistency equation we only need to take into account the linear contribution of the function @xmath315 with respect to @xmath316 .
then , using again the smallness of the variation of the effective transition probabilities and the density of states with respect to changes of energy over intervals of length @xmath7 , @xmath315 can be approximated as @xmath317 using this expression the self - consistency equation takes the form ( [ sc5 ] ) , where the coefficients @xmath123 and @xmath124 are given by @xmath318 @xmath319
to calculate the integral ( [ cd1 ] ) , we again use eq.([sc11 ] ) to change the integration variable from @xmath59 to @xmath147 . furthermore , for convinience
, we introduce the parameter @xmath320 .
doing so , we obtain : @xmath321 where @xmath322 is given by : @xmath323 after an integration by parts we obtain : @xmath324 to calculate @xmath322 , eq .
( [ cdos3 ] ) should be supplemented with the condition @xmath325 . to evaluate @xmath326 ,
it is convenient to use the integration contour @xmath327 , or , introducing polar coordinates @xmath328 : @xmath329 substituting eq.([cdos4 ] ) into eq.([cdos3 ] ) , we have : @xmath330 the value of the integral may be estimated by the saddle - point method , looking for the maxima of the expression @xmath331 .
the stationary point equation is : @xmath332 = 0\ , .
\label{cdos6}\ ] ] as @xmath234 , the stationary point is @xmath333 , and the asymptote of eq.([cdos5 ] ) is : @xmath334 on the other hand , if @xmath237 , we have two stationary points @xmath335 , @xmath336 , @xmath337 , and for @xmath338 the following equation may be obtained : @xmath339 the solution for which at small @xmath340 is : @xmath341the asymptotics of @xmath342 turns out to be : @xmath343 ^{-2}\ , .
\label{cdos8}\ ] ]
for energy dependent density of states it turns out to be difficult to obtain explicite expressions for the time dependence of the energy distribution function . here , according to eq.([pi0 ] ) , the time - dependence has to be calculated from the equation : @xmath344 \ , , \label{sp0}\ ] ] where @xmath99 is given by eq .
( [ sc11 ] ) or ( [ sc13 ] ) .
let us try to integrate over @xmath59 , using the saddle - point approximation .
the saddle - point position @xmath345 may be obtained from the equation : @xmath346 where @xmath347 was introduced . the expression for the diffusion propagator in the saddle - point approximation
may be written as : @xmath348 \,,\ , \label{sp2}\ ] ] where : @xmath349 assuming @xmath350 , we have : @xmath351 , @xmath352 and , finally , the diffusion propagator : @xmath353 \ , . \label{sp4a}\ ] ] at a given time @xmath150 the exponent in the above distribution function is maximal at @xmath354 , where @xmath355 is given by the condition @xmath356 , or : @xmath357 expanding the expression ( [ sp4a ] ) around @xmath354 , we have gaussian distribution : @xmath358 \ , , \label{sp4c}\ ] ] with the dispersion : @xmath359 note , that according to eq.([sp4b ] ) , the dispersion is time dependent . if the density of states @xmath360 , and , consequently , @xmath361 and @xmath362 , are strongly varying functions of @xmath128 , the integrals can be further simplified . in this case
we have : @xmath363 ^{-1}-\left [ \frac{dv\left ( \varepsilon _ m\right ) } { d\varepsilon _ m}\right ] ^{-1}\ , , \label{sp5}\ ] ] @xmath364 } { d\varepsilon _ 0}\right ] ^{-1}-\left [ \frac{d\left [ v\left ( \varepsilon _ m\right ) \omega \left ( \varepsilon _ m\right ) \right ] } { d\varepsilon _ m}\right ] ^{-1}\ , , \label{sp6}\ ] ] @xmath365 ^{-1}-\left [ \frac{dv\left ( \varepsilon _ m\right ) \omega \left ( \varepsilon _ m\right ) } { d\varepsilon _ m}\right ] ^{-1}\ , .
\label{sp7}\ ] ] the applicability condition for the saddle - point method is : @xmath366 ^2\ , .
\label{sp8}\ ] ] note , that this condition restricts the applicablity of the saddle - point approximation to times , that are not too large . also , from eqs.([sp4 ] )
one can see , that this approximation is invalid when @xmath128 is close enough to the initial point @xmath367 , here @xmath368 becomes large , and it is not possible to expand over @xmath369 .
so , the initial stage of the evolution , when the entire distribution is concentrated near @xmath370 , has to be investigated separately too .
here we present details on the calculation of the momenta of the distribution function for large times and for the exponential density of states . in our calculation
we take into account that the density of states decays with decreasing energy .
let us change the integration variable in ( [ dlt4 ] ) , @xmath371 , @xmath372 .
as @xmath373 , we have , introducing @xmath374 , @xmath375 : @xmath376 as @xmath377 and @xmath378 .
both inequalities holds within actual region of integration .
we can write now : @xmath379 ^n\frac \partial { \partial z}\exp \left [ -\frac 1{% \bar\omega \rho _ s^2}z\left ( 1+\frac 12z\right ) \right ] \ , .
\label{mm4}\ ] ] something more about the approximation we have chosen : every extra power of @xmath380 means an extra power of @xmath381 , which can be neglected . an extra power of @xmath382 means the multiplier @xmath383 .
all our considerations are valid if @xmath384 only .
therefore , we have the reason to omit extra powers of @xmath385 in the following , but to keep some extra powers of @xmath386 .
thus we have , for example : @xmath387 \nonumber \\ & = & -\frac 1s\ln \rho _ s+\frac{\sqrt{\bar\omega } } s\int\limits_0^\infty dy\,\left ( y+\frac{\ln \left ( \sqrt{\bar\omega } \rho _ s\right ) } { \sqrt{\bar\omega } % \rho _ s}+\frac{\ln y}{\sqrt{\bar\omega } \rho _ s}\right ) \frac d{dy}\exp \left ( -% \frac{y^2}2-\frac y{\sqrt{\bar\omega } \rho _ s}\right ) , \end{aligned}\]][mm5 ] where @xmath388 within three leading orders on the parameter @xmath389 we have : @xmath390 which can be written , using @xmath391 also as : @xmath392 where @xmath198 is the euler s constant . in the time representation
we have : @xmath393 at large @xmath150 , @xmath394 , the first integral , after the substitution : @xmath395 , may be expanded as : @xmath396 e^y=\ln \ln b\sqrt{% \bar\omega } t \\ & & -\frac 1{\ln b\sqrt{\bar\omega } t } \int_c\frac{dy}{2\pi iy}e^y\ln y-\frac 1{% 2\ln ^2b\sqrt{\bar\omega } t}\int_c \frac{dy}{2\pi iy}e^y\ln ^2y-\dots \\ & = & \ln \ln b\sqrt{\bar\omega } t+\frac
\gamma { \ln b\sqrt{\bar\omega } t}+\frac{\pi ^2/6-\gamma ^2}{2\ln ^2b\sqrt{\bar\omega } t}\,.\end{aligned}\ ] ] as for the second integral , one may restrict oneself with the leading term only . as a result
, we have : @xmath397 let us consider now the second momentum : @xmath398 again , we change the integration variable to @xmath399 , write @xmath373 as @xmath400 , and set the lower limit of the integration to be zero . restricting ourselves with the lowest order in @xmath385 , we set @xmath401 , and : @xmath402 \,.\ ] ] taking into account : @xmath403 we can write : @xmath404 then we have : @xmath405 \\ & = & -\frac { \bar\omega } s\int\limits_0^\infty dy\,\left ( y+ \frac{\ln \left ( \sqrt{\bar\omega } \rho _ s\right ) } { \sqrt{\bar\omega } \rho _ s}+\frac{\ln y}{\sqrt{\bar\omega } % \rho _ s}\right ) ^2\exp \left ( -\frac{y^2}2-\frac y{\sqrt{\bar\omega } \rho _ s}% \right ) \,.\end{aligned}\ ] ] expanding this expressions in power series on @xmath389 , and keeping three leading powers , it may be easily obtained : @xmath406 \frac 1{s\rho _ s^2}\,.\end{aligned}\ ] ] > from eq .
( [ mm6 ] ) we have : @xmath407 so , from eq .
( [ mm10 ] ) we obtain : @xmath408 however , we want to calculate _
central _ momentum : @xmath409 or , in the laplace representation : @xmath410 taking the expression for @xmath411 from eq .
( [ mm7 ] ) , we have : @xmath412 let us consider the first integral .
after changing the integration variable , @xmath413 , we have : @xmath414 closing the integration contour to the left , one can easily prove , that the first intehral in the above expression is zero .
the second one at small @xmath59 may be expanded into powers of @xmath415 . in the leading order
we have : @xmath416 performing integration by part in the integral , and then deforming the integration contour to make it run along the lower and upper shores of the cut @xmath417 of the integration function : @xmath418 \ln x}{(1+x)^2}=% \frac{\pi ^2}6\,.\end{aligned}\ ] ] the evaluation of the second integral in eq .
( [ mm14 ] ) gives : @xmath419 as a result , one can write : @xmath420 thus , one have the second central momentum in the laplace representation to be : @xmath421 and the time dependence of the second central momentum : @xmath422 expression for the third central momentum @xmath423 ( [ dlt7 ] ) may be obtained quite analogously .
one can write the large times distribution function ( [ dlt3 ] ) , after the replacement of the integration variable @xmath424as follows : @xmath425 @xmath426 .
\label{lt1}\ ] ] one can try to calculate the latter integral within the saddle point approximation .
the saddle point equation is : @xmath427 assuming @xmath428 ( remind that @xmath429 ) , we can neglect the third term in the above equation , and to obtain : @xmath430 as we shall see later , the actual values of parameters @xmath174 , @xmath150 , corresponding to the body of the distribution function , obeys the inequality @xmath431 , and , therefore , @xmath432 . but this leads to the saddle point parameter , @xmath433 ^{3/2}\right| = \left| ( z_c+3)/(z_c+2)^{3/2}\right| \approx 1 $ ] , to be of the order of 1 , instead of being much less .
let us note , however , that the function in the exponent in eq .
( [ lt1 ] ) may be represented as a sum of two terms , of which the first one , @xmath434 , ensures the function in the integral to have a maximum at @xmath435 of the width @xmath436 , while the second one , @xmath437 , may be supposed to be almost constant within this peak , it varies essentially at scales @xmath438 within the actual interval of system s parameters .
thus , one can replace that second term in the exponent by its value at some @xmath439 , @xmath440 .
after this , the integration may be easily performed , returning back to the initial integration variable @xmath59 : @xmath441 \right\ } \int\limits_{-i\infty + \delta } ^{+i\infty + \delta } \frac{ds}{2\pi is^2}% e^{st}\,,\ ] ] where the last integral is equal to simply @xmath150 , and , taking into account @xmath442 , @xmath443 , where @xmath444 will be obtained later . performing differentiation with respect to @xmath445 ,
( [ lt1 ] ) , and neglecting @xmath446 , we obtain finally : @xmath447 \right\ } \ , .
\label{lt2}\ ] ] to obtain @xmath448 , let us calculate @xmath449 with the distribution function ( [ lt2 ] ) .
we have : @xmath450\ ] ] @xmath451 where the replacement of the integration variable @xmath452 was done .
assuming @xmath373 , one can write , using @xmath453 as @xmath454 , that @xmath455 ( the condition @xmath456 was used ) . in the leading order on @xmath381
, one can replace @xmath457 within the exponent with @xmath213 , and @xmath458 .
the limits of integration by @xmath459 are to be set @xmath460 .
then we have : @xmath461 \ , .
\label{lt3}\ ] ] this latter integral , analogously to one in eq .
( [ mm5 ] ) , may be evaluated as an expansion on small parameter @xmath462 : @xmath463 comparing eqs.([lt4 ] ) and ( [ dlt5 ] ) , one can see , that @xmath464 .
+ r. stachowitz , phd - thesis _ thermalisation und rekombination optisch angeregter ladungstrger in amorphen silizium _ , marburg ( 1997 ) .
z. ovadyahu and m. pollak , phys .
lett . * 79 * , 459 ( 1997 ) .
m. pollak and z. ovadyahu , j. phys .
i france * 7 * , 1595 ( 1997 ) .
clare c. yu , phys .
lett . * 82 * , 4074 ( 1999 ) .
l. e. golub , s. v. ivanov , e. l. ivchenko , t. v. shubina , a. a. toropov , j. p. bergman , g. r. pozina , b. monemar , and m. willander , phys .
b * 205 * , 203 ( 1998 ) .
e. l. ivchenko , a. n. reznitsky , phil . mag . * 65 * , 733 ( 1992 ) .
l. e. golub , e. l. ivchenko , and a. a. kiselev , j. opt .
b * 13 * , 1199 ( 1996 ) . | a new method for investigating relaxation phenomena for charge carriers hopping between localized tail states has been developed .
it allows us to consider both charge and energy _ dispersive _ transport .
the method is based on the idea of quasi - elasticity : the typical energy loss during a hop is much less than all other characteristic energies .
we have investigated two models with different density of states energy dependencies with our method . in general
, we have found that the motion of a packet in energy space is affected by two competing tendencies .
first , there is a packet broadening , i.e. the dispersive energy transport .
second , there is a narrowing of the packet , if the density of states is depleting with decreasing energy .
it is the interplay of these two tendencies that determines the overall evolution . if the density of states is constant , only broadening exists . in this case
a packet in energy space evolves into gaussian one , moving with constant drift velocity and mean square deviation increasing linearly in time .
if the density of states depletes exponentially with decreasing energy , the motion of the packet tremendously slows down with time . for large times
the mean square deviation of the packet becomes constant , so that the motion of the packet is `` soliton - like '' . 2 |
a spontaneous breakdown of the global @xmath1peccei - quinn symmetry occurs in the course of explaining the absence of charge - parity violation in the theory of strong interactions .
the emerging nambu - goldstone boson known as the qcd axion @xcite constitutes the flag representative for the axion models @xcite and the class of axion - like particles ( alps ) that are predicted in conformal scenarios @xcite , string theory @xcite as well as various standard model extensions , where they are linked to dark matter @xcite . despite experimental efforts toward their detection
much of them exploiting their coherent oscillations into photons mediated by a static magnetic field
there is no evidence yet of alps .
this fact manifests that their interactions with the well established standard model branch might be extremely weak , and the absence of positive detection signals can be used to constrict the associated parameter space .
stringent upper bounds on the alp - diphoton coupling @xmath2 have been inferred from astrophysical considerations .
a plausible generation of alps in the core of stars via the primakoff process might lead to an energy loss which accelerates their cooling and , therefore , their lifetimes .
the nonobservation of a diminishing in the number of stars in the helium - burning phase [ horizontal - branch ( hb ) stars ] within globular clusters constraints @xmath2 to lie below @xmath3 for alp masses @xmath4 below the @xmath5 scale @xcite .
furthermore , as these particle candidates may escape from the sun almost freely , solar alps would likely arrive to earth . by monitoring these hypothetical alp fluxes ,
the cast collaboration has obtained @xmath6 , whenever @xmath4 is below @xmath7 @xcite .
bounds resulting from laboratory experiments are considerably less stringent but free from the uncertainty associated with the underlying astrophysical models @xcite .
some of them have been established from the search of light shining through a wall ( lsw ) @xcite and from scenarios oriented to detect the magnetically - induced vacuum dichroism and birefringence mediated by real and virtual alps , respectively @xcite . while in lsw setups ,
the best upper limit is held by the oscar collaboration @xmath8 @xcite , the current best bound resulting from polarimetric studies has been established by pvlas @xmath9 @xcite .
these limits apply for @xmath10 , and relax significantly for masses larger than @xmath11 by several orders of magnitude . as a general feature both , lsw and polarimetric setups ,
might improve their respective bounds by increasing the field strength and the distance over which it extends . at present , the largest magnetic field generated by superconducting dipole magnets amounts to @xmath12 over a length smaller than @xmath13 .
the incorporation of interferometric cavities for the probe beam allows for extending the interaction region upto five orders of magnitude , but its use has not been enough to push down the bounds in regions of masses @xmath14 , where they turn out to be much less stringent . higher field strengths @xmath15 can be obtained nowadays within the focal spots of high - intensity laser pulses .
even larger magnetic fields @xmath16 , i.e. two orders of magnitude below the critical scale @xmath17 of quantum electrodynamics ( qed ) , are envisaged in the near future within the eli and xcels projects @xcite . despite the inhomogeneous nature of these pulses
confined to short spatial extensions @xmath18their use may allow for the realization of various elusive qed processes @xcite .
for instance , the hibef collaboration @xcite has put forward a laser - based experiment with which vacuum birefringence @xcite may soon be detected .
this kind of experiment provides propitious arenas to test the frontier of the standard model at low energies @xcite and therefore , complements those setups driven by particle accelerators . various theoretical studies in this direction have been carried out to estimate whether high - intensity laser pulses are feasible in the searches for alps @xcite , minicharged particles and hidden photons @xcite .
however , the complicated nature of pulsed laser fields makes the phenomenological descriptions rather challenging and full characterizations of these problems are far from being complete . in this letter
, we provide a step toward the understanding of the role that the pulse profile may play in the search for alps , this way extending a recent investigation carried out within the context of minicharged particles @xcite .
we show that , when dealing with a polarimetric probe driven by the field of a high - intensity linearly polarized pulse , the upper limits resulting from the ellipticity are almost insensitive to the pulse shape , whereas those arising from the rotation of the polarization plane turn out to be more dependent on the field profile . besides
, we reveal that this kind of setup might notably improve the existing laboratory limits in some regions of the alp parameter space .
our investigation relies on the polarization tensor induced by the quantum fluctuations of the axion field over a plane - wave background .
searches for axion dark matter rely on the existence of an alps background permeating the universe . in the following
we will assume that the effects resulting from this nontrivial expectation value are negligible in comparison with the quantum fluctuations that are induced by a pseudoscalar field @xmath19 on the propagation of a small - amplitude electromagnetic wave @xmath20 .
we are interested in evaluating these effects in an external electromagnetic field characterized by the tensor @xmath21 with @xmath22 denoting its four - potential . as long as the fields of interest are minimally coupled , preserving
the formal invariance properties of qed , the relevant equations of motion are are used .
besides , the metric tensor @xmath23 is taken with signature @xmath24 so that @xmath25 . ]
@xmath26 provided that @xmath20 is chosen in the lorenz gauge @xmath27 and @xmath28 . here
the coupling constant @xmath2 and mass @xmath4 are unknown parameters , @xmath29 , whereas the dual of the external field tensor is @xmath30 .
when solving the second equation involved in ( [ dse ] ) and substituting the expression for @xmath31 into the equation associated with the small - amplitude wave , we end up with @xmath32\tilde{\mathscr{f}}^{\sigma\nu}(\tilde{x } ) , \end{split}\label{photoneqmotion}\]]where @xmath33 denotes the alp propagator .
note that eq .
( [ photoneqmotion ] ) has been written in a way that resembles the effective equation of motion of the electromagnetic field in qed , i.e. , including the photon radiative correction .
this fact allows us to identify straightforwardly the polarization tensor @xmath34 induced by the quantum vacuum fluctuations of the pseudoscalar field @xmath19 . the feynman diagram associated with this tensor
is depicted in fig .
[ fig.001 ] .
diagrammatic representation of the vacuum polarization tensor mediated by a quantum fluctuation of a pseudoscalar field @xmath19 in a high - intensity laser pulse [ vertical wavy lines ] . here
, the dashed line represents the alp propagator @xmath35 , whereas the horizontal wavy lines must be understood as the amputated photon legs .
, width=264 ] hereafter , we indentify the background described by @xmath36 with the field produced by a high - intensity laser , the potential of which is taken in the form @xmath37 where @xmath38 are two orthogonal amplitude vectors [ @xmath39 and @xmath40 arbitrary functions of @xmath41 satisfying the boundary conditions @xmath42 . here
, the prime denotes the derivative with respect to the argument , @xmath43 .
the external potential is chosen in the lorenz gauge @xmath44 so that the wave four - vector @xmath45 with @xmath46 and the amplitude vectors @xmath47 satisfy the constraints @xmath48 .
the presence of this plane - wave background allows for introducing the four - vectors @xmath49 which are built up from the amplitudes of the external field modes @xmath50 [ @xmath51 .
these vectors are transverse @xmath52 , orthonormalized according to @xmath53 and satisfy the relation @xmath54 with @xmath55 . here the antisymmetric tensor @xmath56 with @xmath57 is used . in the following ,
we fourier transform eq .
( [ photoneqmotion ] ) and seek the solutions of the resulting equation of motion in the form of a superposition of transverse waves @xmath58 .
correspondingly , @xmath59 in obtaining the expression above we have used the symmetry property @xmath60 . from now on ,
we choose the reference frame in such a way that the direction of propagation of our external plane wave [ see eq .
( [ extfield ] ) ] is along the positive direction of the third axis . as a consequence ,
the external field only depends on @xmath61 via @xmath62 with @xmath63 .
next , we insert the polarization tensor in position space [ see eq .
( [ photoneqmotion ] ) ] into eq .
( [ fourier ] ) .
afterwards , integrations by parts over @xmath64 and @xmath65 are carried out considering the boundary condition @xmath66 .
later , we introduce the light - cone variables @xmath67 , @xmath68 and @xmath69 , @xmath70 .
their use allows us to integrate six out of the eight variables involved in @xmath71 . as a consequence ,
@xmath72where the notation @xmath73 has been introduced .
the tensorial structure of @xmath74 resembles the one associated with the polarization tensor of qed @xcite : @xmath75as @xmath76 , this decomposition does not depend on which choice of @xmath77 is taken ; see also eq .
( [ vectorialbasisbaier ] ) .
the involved form factors @xmath78 depend on the phase of the external field @xmath79 , @xmath80 and @xmath81 .
explicitly , @xmath82where @xmath83 and the change of variable @xmath84 has been carried out . in eq .
( [ formfactors ] ) , the exchange @xmath85 must be carried out only on the index of the field profile functions @xmath86 and on the peak intensity associated with each external field mode @xmath87 with @xmath88 . here
@xmath89 is the fourier transform of @xmath90 .
we remark that also other representations for @xmath91 can be found .
we will see , however , that the chosen one turns out to be convenient for the purposes of this work .
we solve eq .
( [ photoneqmotion ] ) by following a procedure similar to the one used in the context of minicharged particles [ see ref .
if the alp effects do not modify the maxwell equations dramatically , one can solve eq .
( [ fourier ] ) perturbatively by setting @xmath92 . in the following ,
we suppose a head - on collision between the strong laser pulse and the probe beam characterized by a four - momentum @xmath93 , so that @xmath94 and @xmath95 . in accordance , the leading order term is @xmath96 , corresponding to @xmath97 with @xmath98 and @xmath99 the amplitude of mode-@xmath100 .
besides , it follows from eq .
( [ fourier ] ) that the perturbative contribution is given by @xmath101}\sum_ja_{0j}\int d\tilde{\varphi}\\ & \qquad\qquad\times\lambda_{j}^\mu(k)\mathpzc{p}_{\mu\nu}(\tilde{\varphi},k , q_2 ) \lambda^\nu_i(q_2)e^{\frac{iq_{2+}}{\varkappa_+}\tilde{\varphi } } , \end{split}\]]where it must be understood that the only nonvanishing light - cone component of the four - vector @xmath102 is @xmath103 . besides , the poles in the function @xmath104 have been shifted infinitesimally into the complex plane by an @xmath105-term so that correct boundary conditions of the fields at asymptotic times @xmath106 are implemented .
when fourier transforming back , the solution of our problem reads @xmath107 [ see above eq .
( [ fourier ] ) ] , where @xmath108 .
\end{split}\]]here , @xmath109 , @xmath110 , whereas @xmath111 and @xmath112 .
we remark that , in our reference frame , the transversality condition @xmath48 [ see below eq .
( [ extfield ] ) ] implies that @xmath113 .
it can be verified that such a constraint implies @xmath114 to be independent of @xmath115 .
this means that , in the expression above @xmath116 .
structurally , eq . ( [ intermediateequation ] ) coincides with eq .
( 8) found in ref .
this fact allows us to integrate out @xmath117 by using the procedure explained there . as a consequence of this assessment
, the integration over @xmath79 turns out to be restricted to the kinematically allowed region @xmath118 $ ] and we end up with @xmath119 the expression above constitutes the starting point for further considerations .
it holds for arbitrary plane - wave profiles , which formally implies that the pulsed field is infinitely extended in the plane perpendicular to the propagation direction .
this means that , in our model , alps do not experience transverse focusing effects . in an actual experimental realization
, this condition can be considered as satisfied whenever the alp compton wavelength @xmath120 turns out to be much smaller than the characteristic spatial scale , set by the waist size of the pulse @xmath121 . in order words , the outcomes resulting from eq .
( [ degradeddse ] ) are expected to be trustworthy for alp masses @xmath122 . when the external plane wave [ see eq .
( [ extfield ] ) ] is linearly polarized with @xmath123 , @xmath124 , the probe modes in eq .
( [ degradeddse ] ) disentangle from each other .
consequently , we can write the electric field of the probe [ @xmath125 with @xmath126 as a superposition of waves @xmath127\\ & \qquad+\varepsilon_0\sin(\vartheta_0)\pmb{\lambda}_{2}\mathrm{re } \exp\left[-i\phi+\frac{i}{2\varkappa_+k_-}\int_{-\infty}^\varphi d\tilde{\varphi}\ c_4(\tilde{\varphi})\right ] .
\\ \end{split}\label{electricfield}\end{aligned}\]]where the approximation @xmath128 has been used . in the expression above only the leading term , which does not vanish at asymptotically large spacetime distances [ @xmath129 , when the high - intensity laser field is turned off , has been considered .
here @xmath130 refers to the initial electric field amplitude , @xmath131 , whereas @xmath132 is the corresponding initial polarization angle of the probe with respect to @xmath133 , i.e. , the polarization axis of the external pulse . in the expression above , the form factor @xmath134
[ see eq .
( [ formfactors ] ) ] must be evaluated with @xmath135 .
therefore , @xmath136where the relation @xmath137 , with @xmath138 refering to the cauchy principal value , has been applied and @xmath139 denotes the resonant parameter . besides , in this external field configuration , the total probability that a photon with polarization @xmath140 does not decay inside the laser pulse is obtained by evaluating the square of the wave function , @xmath141 . here , @xmath142 refers to the probability that a photon oscillates into an alp , a phenomenon which damps the intensity of the probe beam @xmath143 $ ] as it propagates in the pulse .
the factor responsible for the damping is @xmath144 , provided @xmath145 .
therefore , the vacuum behaves like a dichroic medium , inducing a rotation of the probe polarization from the initial angle @xmath146 to @xmath147 , where @xmath148 is expected to be tiny . at asymptotically large spacetime distances [ @xmath129
, we find @xmath149 as the phase difference between the two propagating modes , @xmath150 , does not vanish either , the vacuum is also predicted to be birefringent . hence , when the strong field is turned off [ @xmath129 , the outgoing probe should be elliptically polarized and its ellipticity is given by @xcite @xmath151 we remark that the last formula is a good approximation only when its right - hand side is smaller than unity .
manisfestly , the relation between eq .
( [ rot ] ) and ( [ elip ] ) is of kramers - kronig type @xcite .
there exists already significant progress in high - purity polarimetric techniques for x - ray probes @xcite which are expected to be exploited in the envisaged experiment at hibef @xcite . in first instance , the planned polarimeter would be designed to measure vacuum birefringence only .
it will involve an analyzer set at a right angle to the initial polarization direction .
this is justified because , in a pure qed context with an x - ray probe and an optical strong field , the rotation of the incoming polarization plane is exponentially small [ @xmath152 for practical purposes ] . in accordance ,
the transmission probability is determined by the ellipticity induced by qed vacuum fluctuations @xmath153 only .
we remark that isolated detections of both @xmath154 and @xmath155 can be carried out if the analyzer is set in such a way that the number of counted photons is minimum .
the axis of the analyzer would form an angle @xmath156 with respect to the initial polarization plane whose measurement allows for establishing @xmath154 . if the minimum count rate differs from zero , this would imply that the outgoing probe beam is elliptically polarized .
therefore , the polarization state transmitted by the analyzer reads @xmath157 .
as a consequence , the transmission probability is @xmath158 ^ 2 $ ] . assuming that @xmath159 , we find that the number of photons to be counted reads @xmath160,\]]where @xmath161 counts the number of laser shots used for a measurement
, @xmath162 denotes the transmission coefficient of all optical components and @xmath163 is the number of probe photons emitted in each shot .
this expression shows that the number of signal photon increases the greater @xmath164 is .
in this section we particularize the optical obsevables given in eqs .
( [ rot ] ) and ( [ elip ] ) for the cases in which the external field is characterized by a gaussian or a @xmath0 envelope [ see fig . [ fig.002 ] ] .
later , in sec . [ genrat ] , we generalize the results found for these two pulses by considering a generic pulse envelope and a carrier envelope phase ( cep ) .
laser pulses with gaussian and @xmath0-profiles are shown in green and red with @xmath165 , respectively .
while the pure gaussian envelope is dotted in blue , the one associated with the @xmath0 function is dashed in black .
these envelopes are given as references .
as the @xmath166pulse falls off faster than the gaussian pulse , its peak strength must be higher ir order to contain the same total energy.,width=264 ] we wish to investigate the optical observables induced by alps when the field profile function is of the form @xmath167\sin(\varphi).\label{fieldprofilefunction}\]]here @xmath168 is defined from the full width at half maximum ( fwhm ) of the field with @xmath169 referring to the number of oscillation cycles within the gaussian envelope .
the fourier transform of this pulse allows us to express the rotation angle [ see eq .
( [ rot ] ) ] in the following form @xmath170assuming the condition @xmath171 , one can use the approximation @xmath172 $ ] .
its substitution into eq .
( [ rotsmalchi ] ) leads to @xmath173since the rotation angle is maximized in the vicinity of @xmath174 , we can anticipate that the most stringent bound will arise at the resonant mass @xmath175 . on the other hand ,
since @xmath176 implies @xmath177 [ see eq .
( [ rotsmalchi ] ) ] , we find that @xmath178 is exponentially suppressed , which indicates that in this regime vacuum dichroism tends to vanish . the ellipticity [ see eq .
( [ elip ] ) ] can be determined straightforwardly by noting that the hilbert transform of a gaussian function @xmath179 is specified by the dawson integral @xmath180 @xcite . with this detail in mind , we find @xmath181\right)\\&\qquad\quad\quad-\mathpzc{d}_f\left(\delta\varphi\left[1-\mathpzc{n}_*\right]\right)-2 e^{-\delta\varphi^2}\mathpzc{d}_f(\delta\varphi \mathpzc{n}_*)\big\vert .
\end{split}\]]let us consider the asymptotic behavior of @xmath155 as @xmath182 .
then , @xmath183 , and @xmath184 .
\end{split}\ ] ] if @xmath185 , we can use @xmath186 and find that @xmath187 $ ] . turning our attention to the situation in which @xmath188 , the resulting expression for the ellipticity in this limit reads @xmath189where
the defining equation of the dawson function has been used , @xmath190 . the last factor in eq .
( [ ndeltaphsmaller1 ] ) , i.e. @xmath191 as @xmath192 .
conversely , the case @xmath193 can be investigated through @xmath194 , in which case the induced ellipticity reads @xmath195 $ ] .
the behavior of the ellipticity in a vicinity of the resonance @xmath196 is also of interest .
in such a case , eq . ( [ elipfinal ] ) reduces to @xmath197for @xmath198 , it leads to @xmath199 $ ] . in order to evaluate the extent to which the projected bounds might depend on the pulse profile
, we will now consider the case in which the function @xmath200 is of the form @xmath201\sin(\varphi)\\ & = \frac{1}{2}\mathcal{r}\sin(\varphi)-\frac{1}{4}\mathcal{r}\sin(\lambda_+\varphi)-\frac{1}{4}\mathcal{r}\sin(\lambda_-\varphi ) \end{split } \label{fieldprofilefunctionsin}\]]for @xmath202 $ ] and zero otherwise .
as before , @xmath203 denotes the number of oscillation cycles within the @xmath166envelope and @xmath204 .
the scaling parameter @xmath205 is chosen in such a way that the total energy of the pulse coincides with the one of the gaussian pulse .
the use of the fourier transform of eq .
( [ fieldprofilefunctionsin ] ) allows us to express the rotation angle [ see eq .
( [ rot ] ) ] in the following form @xmath206 ^ 2 .
\end{split}\]]this expression is characterized by three resonances : @xmath207 and @xmath208 .
while the former is already known from the analysis of the previous case [ see below eq .
( [ imdeltafinal ] ) ] , the remaining ones define two additional resonant masses @xmath209 which do not emerge in the framework of the gaussian pulse .
these extra resonances are direct consequences of the side - band terms arising in the spectral decomposition of the @xmath166pulse , i.e. the last two contributions in the second line in eq .
( [ fieldprofilefunctionsin ] ) .
we note that the behavior of @xmath154 when @xmath210 or @xmath211 , i.e. , in a vicinity of @xmath212 and @xmath213 is given , respectively , by @xmath214comparing the first line of this result with the outcome resulting from eq .
( [ imdeltafinal ] ) we find that the projected sensitivity expected from a @xmath166pulse will be smaller than the one corresponding to a gaussian profile by a factor @xmath215 , approximately .
observe that eq .
( [ rotsmalchisin ] ) tends to vanish as @xmath216 and @xmath217 .
hence , far from the resonant mass @xmath212 the projected bounds to be determined from the rotation angle are expected to be less stringent .
now we focus on the ellipticity [ see eq .
( [ elip ] ) ] . in this case
, it is convenient to use the hilbert transforms @xmath218 and @xmath219 with which we obtain @xmath220 ^ 2\right\vert .
\end{split}\]]observe that @xmath221 as @xmath216 and @xmath217 , whereas for @xmath222 we obtain @xmath223we point out that , for @xmath224 , the expression above coincides with the corresponding outcome resulting from a gaussian pulse [ see below eq .
( [ elipfinalres ] ) ] .
the laser pulses discussed in the previous sections can be understood as particular cases of a more general situation in which the profile function is given by @xmath226here , @xmath225 with @xmath227 and @xmath228 is a real analytic function in the upper half plane which is maximized at @xmath229 , whereas @xmath230 is the cep .
as before , the scaling parameter : @xmath231 guarantees that the pulse energy is invariant with respect to the chosen profile .
moreover , it must be understood that @xmath232 does not depend on time and an explicit evaluation of it can be done by setting @xmath233 . in this context , the angle rotated by the polarization plane [ see eq .
( [ rot ] ) ] @xmath234 coincides with an outcome obtained in ref .
@xcite upto the scaling factor @xmath232 .
there , this formula was established by computing @xmath235 via the s - matrix element associated with the photon - alp oscillations and by exploiting the relation between this quantity and the absorption coefficient [ see discussion above eq .
( [ rot ] ) ] .
the equivalence between both procedures is expected because the optical theorem establishes that @xmath235 is determined by the imaginary part of the vacuum polarization tensor depicted in fig .
[ fig.001 ] . we should however indicate that , in the aforementioned reference , the corresponding expression for the ellipticity was not derived . according to eq .
( [ elip ] ) , this observable reads @xmath236 in these formulae , @xmath237 is the fourier transform of the shape function @xmath225 . the presence of @xmath238 in eqs .
( [ rotcep ] ) and ( [ elipcep ] ) manifests that the projected exclusion regions will depend on the envelope function of the external laser pulse .
besides , these formulae indicate that the cep allows for interference between the @xmath239 and @xmath240 terms whenever @xmath241 and @xmath242
. this interference effect may be constructive or destructive , depending on the overall sign of the associated contribution .
therefore , an appropriate choice of the cep may help to optimize the optical signals .
for the pulses analyzed previously [ see eqs .
( [ fieldprofilefunction ] ) and ( [ fieldprofilefunctionsin ] ) ] , this occurs for @xmath243 with @xmath244 .
first we estimate the projected limits considering the benchmark parameters of the proposed experiment at hibef @xcite . in this setup the strong field will be produced by a petawatt laser operating in the optical regime with @xmath245 [ @xmath246 , a repetition rate of @xmath247 , a temporal pulse length of about @xmath248 [ @xmath249 , and a peak intensity @xmath250 .
the envisaged probe beam is the european x - ray free electron laser , operating with frequency @xmath251 and delivering @xmath252 photons per shot .
the transmission coefficient of the optics to be used in this experiment is @xmath253 , and the incoming polarization angle will be @xmath254 .
under such a conditions , a qed signal as small as @xmath255 is likely to be reached provided a perfect overlapping between the probe and the strong laser field is achieved @xcite .
an exclusion region can be inferred from this projected result by assuming that the induced ellipticity due to alps does not overpass the upper bound set by @xmath256 .
it is shaded in green in the right upper corners in fig .
[ fig : mb003 ] . while the outcome shown in the left panel relies on the gaussian model , the one in the right panel
is based on the @xmath166pulse . in each panel
, there is a tiny wedge shaded in red , which is ruled out by supposing that the rotation angle can be measured with a sensitivity of the same order of magnitude of @xmath256 .
our estimates reveal that the most stringent bounds @xmath257 [ gaussian profile ] and @xmath258 [ @xmath166profile ] would emerge at the resonant mass @xmath259 [ see below eq .
( [ imdeltafinal ] ) ] .
as we already pointed out , the energy scale associated with the waist size of the pulse @xmath121 limits the validity of our predictions towards smaller alp masses . in view of the strong
focusing applied at hibef [ @xmath260 , our potential discovery applies for @xmath261
. we must , in addition , mention that this result relies on the forward scattering analysis , and that the waist size of the probe @xmath262 is bigger than @xmath121 .
this situation in combination with the nonconservation of the transverse momentum that the focusing induces , is favourable to scatter probe photons slightly off the forward direction .
such an effect has been proposed as an alternative way to detect the qed birefringence at a small angle @xcite .
it might as well be beneficial for the search of alps as the signal - to - noise ratio for photons transmitted through the analyzer improves notably .
however , this study would require to incorporate the focusing effects , which is still beyond the scope of this work .
if the planned hibef experiment was driven by the strong field to be reached at eli [ @xmath263 , @xmath245 , @xmath264 , corresponding to @xmath265 , and the sensitivity remained within the same order of magnitude @xmath266 , the limits above would be pushed down to @xmath267 [ gaussian model ] and @xmath268 [ @xmath166pulse ] at the resonant mass @xmath259 .
the projected areas to be excluded from the ellipticity [ rotation angle ] can be seen in fig .
[ fig : mb003 ] in black [ brown ] .
we emphasize that the pulse at eli is expected to be strongly focused [ @xmath269 .
hence , the exclusion areas found for this setup are expected to be trustworthy as long as @xmath270 .
our estimates reveal that the shape of the bounds resulting from the ellipticity almost coincide for both pulse models . however , the borders of the excluded areas coming from the rotation angle differ from each other more strongly .
the described behavior is even more pronounced when both observables are probed with an optical laser beam and the envisaged eli laser drives the vacuum polarization .
the potential exclusion regions associated with this case are summarized in fig .
[ fig : mb003 ] in blue [ ellipticity ] and purple [ rotation angle ] .
they have been found by supposing a sensitivity of the order of @xmath271 , a probe frequency @xmath272 and a probe intensity much smaller than the one of the strong laser field .
we remark that the outcomes resulting from this hypothetical setup were obtained by considering a counterpropagating geometry @xmath273 and an initial polarization angle @xmath254 . to conclude , we study a situation in which the strong field is generated by the nanosecond front - end of the phelix laser @xcite , [ @xmath274 , @xmath275 , @xmath276 , @xmath277 , corresponding to @xmath278 .
it is worth mentionig that the electromagnetic pulse produced by this system closely approaches to a monochromatic plane wave as the conditions @xmath279 and @xmath280 are fulfilled simultaneously .
the potential exclusion bounds which follow from this setup are valid whenever the condition @xmath281 is satisfied . as in the previous case , we take the probe beam with a doubled frequency @xmath282 . besides , we suppose that its intensity and waist size are much smaller than the corresponding quantities of the strong field . assuming some achievable conditions such as @xmath254 , a counter propagating geometry and a sensitivity @xmath283 , we find that the area shaded in cyan [ fig . [
fig : mb003 ] ] could be excluded potentially .
this projected result shows that large sensitivities can be achieved , provided the number of cycles @xmath169 is large enough to compensate for the relative smallness of the laser intensity @xmath284 .
we have studied the discovery potential that modern and envisaged laser systems offer in the search for alps .
our investigation reveals that laser - based setups , designed to detect the hitherto unobserved qed vacuum birefringence , may provide stringent bounds on the alp - diphoton coupling @xmath2 in mass regions where the constraints resulting from experiments driven by dipole magnets are considerably less severe .
special attention has been paid to the consequences resulting from the pulse profile function , which were evaluated by solving perturbatively the system of equations describing the oscillation of photons into alps mediated by a generic plane - wave background .
our analysis points out that a broad sector of the parameter space of alps might be discarded , no matter what type of strong field profile is utilized .
the precise location of the projected exclusion areas will depend , in general , on the chosen pulse profile and the cep .
however , a direct comparison between two pulses with different envelope has indicated that the outcomes resulting from the ellipticity are less sensitive to this dependence than the bounds arising from a plausible rotation of the polarization plane . besides
, we have pointed out that an appropriate choice of the cep may optimize the alps search . | hypothetical oscillations of probe photons into axion - like particles might be revealed by exploiting the strong fields of high - intensity laser pulses .
considering an arbitrary plane - wave background , we determine the polarization tensor induced by the quantum fluctuations of the axion field and use it to calculate how the polarimetric properties of an initially linear - polarized probe beam are modified .
we find that various experimental setups based on contemporary facilities and instrumentation might lead to new exclusion bounds on the parameter space of these particle candidates .
the impact of the pulse shape on the discovery potential is studied via a comparison between the cases in which the wave is modulated by a gaussian envelope and a @xmath0 profile .
this analysis shows that the upper limits resulting from the ellipticity are relatively insensitive to this change , whereas those arising from the rotation of the polarization plane turn out to be more dependent on the field shape . beyond the standard model , axion - like particles , vacuum polarization , laser fields .
14.80.-j , 12.20.fv |
the jlab 6 gev research program in hadronic physics can be separated into three broad areas of investigation : the structure of the nuclear building blocks ; the structure of nuclei ; and symmetry tests in nuclear physics . here , the results of eight recent experiments are presented that answer detailed questions of two of these areas : * measure precisely the nucleon s charge and magnetization distribution * determine the internal structure of the nucleon in the valence region * develop the experimental methods for performing tomography of the nucleon * probe the nuclear interior with a controlled impurity to learn about deeply - lying shell structure * clarify the short - range nature of nucleon - nucleon interactions in nuclei and compare the properties of bound nucleons with free ones * test chiral perturbation theory by studying the properties of goldstone bosons at present the continuous electron beam accelerator facility ( cebaf ) accelerates electrons to 6 gev by recirculating the beam four times through two superconducting linacs , each producing an energy gain of 600 mev per pass . in the past decade there has been impressive progress at jlab in improving the quality of the polarized beam , such that a beam can be delivered routinely with a polarization of @xmath085% and an intensity of up to 200 @xmath1a . the base instrumentation in hall a has been used for experiments which require high luminosity and high resolution in momentum and/or angle of at least one of the reaction products .
the central elements are the two high resolution spectrometers ( hrs ) , to which recently a third spectrometer has been added with a large acceptance ( bigbite ) . the cebaf large acceptance spectrometer ( clas ) in hall b is used for experiments that require the detection of several , loosely correlated particles in the hadronic final state at a limited luminosity .
the hall c facility has generally been used for experiments which require high luminosity at moderate resolution .
the core spectrometers are the high momentum spectrometer ( hms ) and the short orbit spectrometer ( sos ) .
the hms has a maximum momentum of 7.6 gev/@xmath2 .
of the experiments presented here , only @xmath3he(@xmath4 ) used just the base equipment , all others needed additional specially designed instrumentation and three experiments , primakoff , hypernuclear spectroscopy in hall c and @xmath5 used only the beam line in the hall it was run .
an upgrade of the cebaf accelerator to 12 gev has recently obtained the cd-2 approval , an essential component of the formal doe process for large projects .
the project also includes the construction of a fourth experimental hall and the upgrade of two of the existing halls .
it is on track to provide a first 11 gev beam in 2013 .
although the partonic structure of the proton has been studied in great detail , much less is known about that of the neutron , mainly due to the unavailability of a free neutron target .
thus , the information about the neutron had to be extracted from scattering data on nucleons bound in nuclei which resulted in substantial theoretical uncertainties . a novel detection technique , called bonus - for barely off - shell nucleon structure - has been recently implemented in the clas detector in hall b that allows to measure the inclusive electron scattering off an almost free neutron , eliminating to a large extent nuclear binding effects .
this technique uses a low - momentum recoil detector to tag slow backward - moving spectator protons - with a momentum as small as 70 mev/@xmath2 - in coincidence with the scattered electron in the reaction @xmath6h(@xmath7)x , semi - inclusive deep - inelastic scattering ( sidis ) . ] the structure functions of the nucleon reflect a variety of aspects of qcd , from asymptotic freedom at large momenta down to confinement at the hadronic scale . from measurements of these structure functions in the scaling region
one can extract the parton distribution functions ( pdf ) , the information on the momentum and spin carried by the quarks as a function of bjorken @xmath8 . in the non - perturbative region
the structure functions in the nucleon resonance regime can be compared to qcd - inspired models .
the applicability of quark - hadron duality can be probed by comparing the resonance structure and the partonic structure .
although the high - quality jlab data on @xmath9 and @xmath10 in the resonance region , when averaged over the resonances , show remarkable agreement with the extrapolation of dis results in the perturbative region down to lower @xmath11 at comparable @xmath8 , no such comparison has been possible for the neutron .
another open question of similar impact is the behavior of the structure functions as @xmath8 approaches 1 , in the region where the @xmath12 and @xmath13 valence quarks dominate , when all of the nucleon momentum is carried by a single quark .
simple phenomenological models like su(6 ) symmetry predict a significantly different behavior than perturbative qcd or quark models with improved hyperfine interactions .
figure [ fig : bonuslayout ] shows a schematic lay - out of the detector .
it basically consists of a radial time - projection chamber ( rtpc ) surrounding a thin deuterium target .
the rtpc is relatively small , 20 cm long and 13 cm in diameter , situated at the standard clas target position , inside the superconducting solenoid designed for the dvcs program .
the target is a 20 cm long , 6 mm diameter kapton cylinder , holding deuterium gas at 7.5 atm . the 30 cm long solenoid is used to suppress the background due to mller electrons in both the recoil detector and the forward drift chambers .
the rtpc is designed to provide position and timing information sufficient to identify low - momentum protons in the back hemisphere .
it consists of 3 cylindrical gas electron multiplier ( gem ) detectors - the first implementation of curved gems - mounted around a kapton inner window with read - out pads on the outside .
charged particles leave a trail of electron - ion pairs on passing through the sensitive gas inside the rtpc .
a radial electric field forces the electrons to drift outwards while the field focusing due to the holes in the gems causes avalanche multiplication . finally , the signal is collected on the outer surface by individual pads .
the location of the pads provides position information and the arrival time provides a measure of the radius where the electrons were produced . from the amount of charge
produced the @xmath14 can be estimated and thus the mass of the particle traversing the rtpc .
spectrum reconstructed using the backward - going spectator protons detected in bonus ( grey area ) , compared to a @xmath15 spectrum from dis off the deuteron.[fig : bonuscomm ] ] the measurement of tagged structure functions in sidis off the deuteron , by detecting a slow recoil proton in the back hemisphere , will resolve the ambiguities introduced by nuclear effects in extracting neutron structure functions from dis off the deuteron . by detecting the spectator proton with the smallest momentum feasible ,
one minimizes the degree to which the struck neutron is off - shell and thus the associated uncertainties .
also , at such low momenta the deuteron wave function ( or more accurately the spectral function ) is well known . in the back hemisphere corrections due to target fragmentation
are predicted to be very small .
lastly , final - state interaction ( fsi ) effects have been calculated to contribute less than 5% to the spectral function - and thus to the extraction of @xmath16 - for spectator momenta less than 100 mev/@xmath2 at angles larger than 130@xmath17 .
the bonus detector was successfully installed and commissioned in the fall of 2005 , followed by a first two - month production run for experiment e03 - 012@xcite with a 40 na beam at an energy of 3.2 gev .
the bonus detector operated to full satisfaction at a luminosity of @xmath18 @xmath19s@xmath20 .
the vertex could be reconstructed with an accuracy of @xmath01 cm ( see fig . [
fig : bonuscomm ] ( left ) ) . in the right part of that figure the invariant - mass spectrum reconstructed with the bonus detector ( grey area )
is compared to one obtained from dis off the deuteron .
note that the width of the resonances has decreased thanks to the absence of fermi motion effects in the bonus data .
the data obtained will provide accurate values of the ratio @xmath21 , and thus of @xmath22 , up to an @xmath8-value of @xmath00.5 .
possible modifications of the properties of nucleons by the nuclear medium that they are bound in , have long been the subject of study .
early conclusions drawn from the longitudinal / transverse ratio of the cross section for electron - induced proton knock - out have since been convincingly refuted@xcite . however , several calculations suggest a measurable deviation of the nucleon electro - magnetic form factors ( emff ) from their free - space values already at modest momentum - transfer values @xmath23 2.5 gev@xmath6
. polarization transfer in quasi - elastic nucleon knock - out , ( @xmath24 , e@xmath25 ) , provides the required sensitivity to the properties of the nucleon in the nuclear medium . in free electron - nucleon scattering
the measured ratio of the transverse and longitudinal polarization , transferred to the proton , is directly proportional to the ratio of the charge and magnetic form factors ( see section [ gen ] ) .
this property has been used in a study of polarization transfer in @xmath3he , chosen as a target nucleus because of its high density and relatively simple structure , that facilitates microscopic calculations . as a function of @xmath11 from mainz@xcite and e93 - 049@xcite ( open symbols ) along with preliminary results from e03 - 104 ( filled circles ) : @xmath26 is the ratio of transverse to longitudinal polarization of the knocked - out proton in @xmath3he(@xmath4 ) compared to the same ratio for @xmath27h(@xmath28 ) .
the baseline @xmath29 is the value of @xmath26 obtained in a plane - wave calculation , to account for `` trivial '' effects of free @xmath30 moving protons .
the data are compared to calculations from the madrid group@xcite and schiavilla _
et al._@xcite.[fig : poltrans ] ] earlier @xmath3he(@xmath4 ) polarization - transfer experiments were performed at mainz@xcite at @xmath11 = 0.4 gev@xmath6 and in hall a@xcite at @xmath11 = 0.5 , 1.0 , 1.6 and 2.6 gev@xmath6 . a recent experiment , again in hall a@xcite , has extended these measurements in 2006 with high precision at 0.8 and 1.3 gev@xmath6 .
the data were taken in quasi - elastic kinematics at low missing momentum in order to minimize fsi and meson - exchange corrections .
one of the two high - resolution spectrometers in hall a was used to detect the scattered electron , the other to detect the knocked - out proton .
the azimuthal distribution of protons scattered by a graphite analyzer was measured in a focal - plane polarimeter@xcite to yield the polarization observables .
the two - body character of the reaction was identified with the missing - mass technique .
polarization transfer in the @xmath27h(@xmath28 ) reaction was measured at the same kinematics in order to study the differences between the in - medium polarization transfers and the free values .
the proton form - factor ratio @xmath31 extracted from the hydrogen data was found to be in excellent agreement with previous polarization - transfer data@xcite .
the results of the present experiment are expressed in terms of the polarization - transfer double ratio , in which the @xmath3he polarization - transfer ratio is normalized to the @xmath27h one , measured under identical conditions : @xmath32 nearly all systematic uncertainties cancel in @xmath26 , the remaining small uncertainties are predominantly due to the spin transport in the hrs spectrometer .
the induced proton polarization @xmath33 was also extracted from the helicity - independent fpp data and is a direct measure of fsi effects .
the preliminary results for the double ratio @xmath26 are shown in fig .
[ fig : poltrans ] and are clearly in agreement with the existing data .
the inner bars indicate the statistical errors , the outer ones the linear sum of statistical and systematic errors .
the final systematic errors are expected to be significantly smaller than those shown here .
the relativistic distorted - wave impulse - approximation ( rdwia ) calculations from udias _
et al._@xcite by themselves overpredict the data by @xmath06% . on the other hand , if the density - dependent medium - modified form factors calculated by lu _ et al._@xcite are folded into udias calculations ( qmc ) , a good agreement is obtained with most data , indicating possible modifications of the proton emff by the nuclear medium
. however , the observed suppression of the super - ratio has been equally well described by a more traditional calculation by schiavilla _
et al._@xcite with free form factors in which charge - exchange fsis and meson - exchange and isobar current effects have been included .
the high statistics of the present experiment@xcite will allow a study of the individual polarization observables as a function of the missing momentum @xmath34 .
it should be pointed out that the observed low value of @xmath26 at @xmath11 = 0.8 gev@xmath6 could be interpreted as an indication of a @xmath11-dependence of @xmath26 not predicted by any of the calculations shown . ; note that the experimental data have been corrected for the spectrometer acceptance .
[ fig : inducpol ] ] the results for the induced polarization @xmath33 are shown in fig . [
fig : inducpol ] . as mentioned earlier , @xmath33 is directly sensitive to fsi effects .
the present large size of the systematic errors is due to the uncertainty in instrumental asymmetries that will decrease strongly with further analysis .
the data have been corrected for the hrs acceptance to facilitate a comparison with the calculations .
the data suggest that the fsi effects have been overestimated in schiavilla s calculations
. it will be interesting to see how his calculations for @xmath26 will change once he has optimized his calculations for @xmath33 .
the two - photon decay mode of the neutral pion @xmath35 is completely due to the quantum fluctuations of the quark fields coupling to a gauge field , a symmetry breaking of purely quantum - mechanical origin . in the chiral su(2 )
limit of vanishing quark masses its decay width can be expressed in terms of the pion decay constant @xmath36 and the fine structure constant @xmath37 @xmath38 ^ 3\left[\frac{\alpha}{f_{\pi}}\right]^2 = 7.725 \pm 0.044 \rm{~ev}\ ] ] recently , two calculations@xcite have studied the effect of the explicit breaking of the su(2 ) symmetry induced by the non - zero quark masses in next - to - leading order , resulting in predictions for the decay width of 8.10 @xmath39 0.08 ev and 7.93 @xmath39 0.12 ev .
the main contribution to the shift upwards of nearly 0.4 ev is due to isospin breaking that introduces components from the @xmath40 and @xmath41 in the physical @xmath35 .
decay width from the @xmath42c data of the primex experiment@xcite compared to earlier results from cern@xcite , cornell@xcite , tomsk@xcite and desy@xcite .
the predictions from chiral perturbation theory are also shown , to leading and to next - to - leading order@xcite and one using the qcd sum rule@xcite.[fig : primexres ] ] however , the experimental knowledge of the decay width is much less established , with a current world - average value of 7.74 @xmath39 0.55 ev . in the first place
the assigned error is nearly an order of magnitude larger than that of the theoretical estimate , but the experimental results show an even larger dispersion .
two methods have been used to measure the @xmath35 decay width ( or equivalently its lifetime ) .
a direct measurement can be made by observing the decay distance between the production and decay points , using a two - foil technique , in which the pion is produced in the first foil and the decay photons are pair converted in the second foil .
the only such direct measurement has been performed at the cern sps@xcite .
the alternative method uses the primakoff effect , in which photons produce pions in the coulomb field of a nucleus .
three such measurements , all using bremsstrahlung beams , have been reported@xcite with a large dispersion as is evident from fig . [
fig : primexres ] .
the tagged photon facility in hall b at jlab provides significant advantages for a new high - precision measurement of the @xmath35 decay width using the primakoff effect : the well - defined photon energy enables a clean separation of background processes and the tagging technique allows a better control of systematic errors . besides the primakoff effect , three additional processes contribute to pion photoproduction at high energies : nuclear coherent and incoherent scattering and interference between primakoff and coherent scattering .
these processes can be separated by their different angular behaviors ( see fig . [
fig : primexcross ] ) .
consequently , a detector is needed for the @xmath35 detection with good angular resolution to identify the competing processes and good energy resolution to suppress multi - photon backgrounds through cuts on the invariant mass .
a hybrid design was chosen for this calorimeter ( hycal ) in order to optimize the performance and cost .
the central part contains 1152 pbwo@xmath43 crystals , 20.5 x 20.5 x 180 mm@xmath44 each , arranged in a square array with a four - block hole in the center to allow the beam to pass through .
the central part is surrounded by 576 lead glass blocks , 38.2 x 38.2 x 450 mm@xmath44 each .
the total size of the calorimeter , 116 x 116 cm@xmath6 , is sufficient to detect both decay photons from a @xmath35 meson produced at the target 7.3 m upstream .
an essential requirement for the experiment is to determine the cross section with high absolute accuracy .
the photon flux was measured with a lead - glass total absorption counter that was put into the beam periodically and monitored continuously with a pair spectrometer that detected the e@xmath46e@xmath47 pairs produced in the target using a dipole magnet and a set of scintillators .
the thickness of the two targets used , @xmath42c and @xmath48pb , was measured to an accuracy of 0.04% . cross section on @xmath42c , together with the contributions from the primakofff effect ( dashed ) , from nuclear coherent ( dot - dashed ) and incoherent ( grey full ) scattering and from the interference ( short - dashed ) between primakoff and coherent scattering.[fig : primexcross ] ] the experiment@xcite ran for three months ( one month for commissioning and two for production running ) in late 2004 at a beam energy of 5.7 gev .
the energy and angular resolution of the central part of the calorimeter were shown to be 2.3% ( at a photon energy of 1 gev ) and 0.34 mrad , respectively , resulting in an invariant mass resolution of 2.3 mev/@xmath49 at the @xmath35 mass .
a very clean identification of elastic @xmath35 events was obtained after a final correlated cut on the invariant mass and the so - called elasticity - the ratio of the energy of a cluster pair and the tagger energy - .
the angular distribution of those events , using only the central part of the calorimeter , is shown in fig . [
fig : primexcross ] .
finally , the primakoff cross section was extracted from a fit to the angular distribution in which the amplitudes for the primakoff , coherent and incoherent scattering and the phase of the interference term were varied . the preliminary result for
the @xmath35 decay width from only the @xmath42c data is 7.93 ev @xmath39 2.1% @xmath39 2.0% , in excellent agreement with the nlo predictions .
the main limitations to reach the projected goal of a 1.4% accuracy are statistics ( presently 1.8% , caused by cutting the run short and limiting the analysis so far to the central part of hycal ) and the uncertainty in the photon flux ( presently 1.1 % ) .
in the final analysis both those issues will be addressed .
information about the deeply - lying shell structure of the nuclear interior is difficult to obtain since all available nucleon levels are filled , and the addition of a nucleon to a low - lying shell is blocked by the pauli principle . because a @xmath50 hyperon does not suffer from pauli blocking , it can be inserted deep inside a nucleus as an impurity , thus providing a sensitive probe of the nuclear interior .
in addition , the elementary hyperon - nucleon interaction can be studied with great sensitivity in hypernuclear spectroscopy .
nuclear matter containing strangeness is also predicted to play an important role in stellar objects such as neutron stars . in the past , hypernuclear
spectroscopy has been studied with hadronic reactions , severely hampered by the poor beam quality of the secondary pion and kaon beams .
kaon electro - production from nuclei makes it possible to deposit a tagged @xmath50 deep inside the nucleus , and to observe its interaction with the nuclear system . a excellent energy resolution ( to separate the individual states in the hypernuclear system ) and a high beam intensity ( to compensate for the small cross section )
are essential requirements for such studies .
electro - production provides a number of advantages over earlier studies@xcite : * the excellent beam quality makes it possible to identify hypernuclear states with a resolution of potentially @xmath0300 kev * due to the large momentum transferred and strong spin - flip amplitudes in electro - production hypernuclear states can be created with natural and unnatural parity and low and high spin * the ( @xmath51 ) pair is produced on the proton in contrast to hadronic studies where it is produced on the neutron , making it possible to study different hypernuclei and charge - dependent effects from a comparison of mirror hypernuclei , @xmath52 @xmath53b - @xmath53c the multi - gev high - intensity beam at jlab thus provides a unique opportunity to study hypernuclear spectroscopy .
two different implementations have been chosen in halls a and c. in hall c the so - called `` zero - degree tagging '' configuration was chosen , whereby the production rate was optimized by detecting both the scattered electrons and the kaons at extreme forward angles with a beam energy close to the reaction threshold , @xmath01.8 gev .
a splitter magnet was used to separate the electrons and the kaons to different spectrometers . in a first experiment@xcite a missing - mass resolution of 900 kev was obtained in @xmath53b , but the production rate was severely limited by accidental coincidences from bremsstrahlung . in a follow - up experiment@xcite the electron spectrometer , the same enge split - pole as in the first experiment , was slightly - by @xmath08.5@xmath17 - tilted vertically which significantly suppressed both the bremsstrahlung and the mller rate . also the sos spectrometer ( short - orbit ) was replaced by a new high - resolution kaon spectrometer ( hks ) with two layers of water and three layers of aerogel erenkov detectors for particle identification . in this new configuration data were taken in 2005 with a hypernuclear production rate of @xmath010 per hour .
a preliminary analysis has indicated a missing - mass resolution of @xmath54 500 kev . for the following stage
a new electron spectrometer has been designed and constructed for a tenfold increase in production rate and a energy resolution as small as 300 kev .
this experiment@xcite is expected to run in 2009 .
b. the results of the best fit ( solid curve ) and that of an unnormalized theoretical calculation ( dashed curve ) are shown superimposed on the data
. see text for further details .
right : the same for @xmath55n .
[ fig : hyper - halla ] ] in hall a the pair of high - resolution spectrometers ( hrs ) was used , supplemented with a superconducting septum magnet to permit each hrs to detect particles at angles a small as 6.5@xmath17 . highly selective particle identification
was obtained through the use of a ring - imaging erenkov ( rich ) detector in a proximity - fusing geometry , a 15 mm freon radiator and a csi photocathode , coupled to two aerogel erenkov detectors . because the path length in each hrs is @xmath020 m , kaons had to be produced with a rather large momentum , so a beam energy of 4 gev was selected .
experiment e94 - 107@xcite took data on a @xmath42c target in 2004 and on a waterfall target for hypernuclear production on oxygen in 2005 .
the measured excitation energy ( @xmath56 ) spectrum for @xmath53b@xcite is shown in the left part of fig .
[ fig : hyper - halla ] . after subtracting
the background evaluated from random coincidences in a large timing window , no residual background was observable in the negative @xmath56-range . also shown in fig .
[ fig : hyper - halla ] is a fit to the six regions with an excess of counts above background .
an energy resolution of better than 700 kev fwhm was obtained for the peaks fitted , in large part due to carefully minimizing the primary energy spread of the electron beam - to less than @xmath57 fwhm - and stabilizing the beam energy centroid during the experiment .
theoretical calculations in the distorted wave impulse approximation ( dwia ) using the sla model@xcite for the elementary @xmath58 reaction and shell - model wave functions for @xmath59b , show a very good overall agreement with the data without any normalization .
this is the first time that in a hypernuclear spectrum a measurable strength has been observed in the core - excited part .
the structures at @xmath02.7 and @xmath05.8 mev can be well described by coupling a @xmath50 in a @xmath60 state to a 1/2@xmath47 and a 3/2@xmath47 state in @xmath59b . in the right part of fig .
[ fig : hyper - halla ] the hypernuclear spectrum is shown for @xmath55n@xcite . here , four peak regions could be identified . a similar calculation as for @xmath53b
did reproduce the observed cross section for the four peaks , but with non - negliglible discrepancies in the excitation energies .
the use of a waterfall target provided simultaneous cross - section data for the elementary @xmath58 reaction .
the nucleon - nucleon force is strongly attractive at long range , but at short range the nuclear force changes sign and then becomes highly repulsive .
a survey of proton knock - out by electrons@xcite has shown that the spectroscopic strengths for valence orbitals amounts to only @xmath060% of its sum rule value .
most of this depletion was assigned to nucleon - nucleon short - range and tensor correlations , that have moved the strength to large values of the missing energy @xmath61 and the missing momentum @xmath34 .
a systematic study@xcite of the @xmath42c(@xmath62 ) reaction in parallel kinematics established @xmath011% correlated proton strength at @xmath34-values above the fermi surface , but @xmath63 80 mev , and @xmath012% at @xmath64 80 mev . strongly correlated nucleon pairs are characterized by a large relative momentum , but a small center - of - mass momentum . indeed , the existence of short - range n - n ( and 3n ) correlations has been quantitatively established through the scaling behavior at large energy and momentum transfer , @xmath65 and @xmath11 , as a function of @xmath8 of inclusive electron scattering @xcite . the ratio of the inclusive cross section of a number of nuclei @xmath30 that of @xmath44he was observed to become constant at @xmath66 and again at @xmath67 , clearly indicating scattering off 2(3)-body correlated states . from the values of these ratios
it was established that @xmath020% of all nucleons in @xmath42c occur in correlated pairs ( and @xmath00.2% in three - nucleon configurations ) .
et al._@xcite have criticized this interpretation , claiming that fsi effects were not taken into account . however ,
laget@xcite and frankfurt@xcite have shown that fsi effects are indeed very large from the correlated configurations , but then cancel in the ratio .
c(@xmath68)/@xmath42c(@xmath69 ) ratio@xcite , the extrapolated @xmath42c(@xmath68)/@xmath42c(@xmath70 ) and @xmath42c(@xmath69)/@xmath42c(@xmath70 ) yield ratios and the yield ratio of the @xmath42c(@xmath71)/@xmath42c(@xmath72 ) reaction from ref .
( @xcite ) .
right : the ration of the @xmath73 and the @xmath74 momentum distributions as calculated by schiavilla _
et al._@xcite compared to the ratio of @xmath73 and @xmath74 pairs extracted from the measured @xmath42c(@xmath68)/@xmath42c(@xmath69 ) ratio .
also shown are the projected @xmath34-values and errors for an approved experiment to measure the ratio @xmath75he(@xmath68)/@xmath75he(@xmath69)@xcite .
[ fig : srcresult ] ] experiment e01 - 015@xcite studied simultaneously the @xmath42c(@xmath62 ) , @xmath42c(@xmath76 ) and @xmath42c(@xmath77 ) reactions .
it ran in hall a in 2004 at a beam energy of 4.6 gev .
the two hrs spectrometers were used to identify the @xmath42c(@xmath62 ) reaction .
the kinematic conditions for single - proton knock - out were ( @xmath78 , @xmath65 , @xmath8 ) = ( 1.65 gev/@xmath2 , 0.865 gev , 1.2 ) . knocked - out protons were detected in the right hrs at three @xmath34-values : 0.35 , 0.45 and 0.55 gev/@xmath2 .
e01 - 015 was the first experiment to use the large acceptance spectrometer bigbite , consisting of a large dipole magnet and equipped with a detector package consisting of three planes of plastic scintillators , segmented in the dispersive direction .
positioned at 1.1 m from the target , it provided an angular acceptance of 96 msr and a momentum acceptance from 0.25 to 0.9 gev/@xmath2 . directly behind bigbite ,
a 0.4 m thick neutron detector was positioned at a distance of 6 m from the target .
it contained 88 plastic scintillators , arranged to match the solid angle of bigbite by covering an area of 1 x 3 m@xmath6 .
these unshielded systems operated reliably at a luminosity of @xmath79 @xmath19s@xmath20 .
bigbite was positioned to detect a proton that balanced the @xmath34 for the @xmath42c(@xmath62 ) reaction .
such recoiling protons were identified by requiring an energy loss and a flight time measured with the scintillators to be consistent with the momentum setting of bigbite .
contributions from @xmath80-resonance excitation were removed by angular cuts on @xmath81 .
the observed angular distribution of the coincident proton pairs was in agreement with the correlated pair having a cm motion relative to the @xmath82 spectator system with a width of @xmath0140 mev / c .
as shown in fig .
[ fig : srcresult ] for 9.5 @xmath39 2% of the @xmath42(@xmath62 ) events , constant in the @xmath34-range from 300 to 600 mev/@xmath2 ,
a second proton is ejected roughly back - to - back with the first one@xcite .
estimates of fsi and charge - exchange contributions were obtained using the glauber approximation and were shown to cancel each other to a large extent .
this conclusion is supported by the isotropy of the angular distribution of the coincidences , mentioned earlier . for each @xmath42c(@xmath62 ) event , both the bigbite and the neutron detector were read out , so that proton and neutron data were collected under identical conditions .
the ratio of @xmath83(@xmath77)/@xmath83(@xmath76 ) events was measured to be 9.0 @xmath39 2.5@xcite .
since the experiment can only detect 50% of the initial - state @xmath84 pairs ( those with the proton momentum anti - parallel to the virtual photon ) , the ratio of @xmath84 and @xmath85 pairs is 18 @xmath39 5 .
the ratio @xmath83(@xmath77)/@xmath83(@xmath62 ) measured to be 0.96 @xmath39 0.23 ( fig .
[ fig : srcresult ] ) indicates that in the @xmath34-range of 400 to 600 mev/@xmath2 all protons are in a pair configuration .
recent calculations @xcite have shown that the observed dominance of @xmath84 pairs is a clear fingerprint of the short - range tensor component of the nucleon - nucleon force , independent of the parametrization of the nucleon - nucleon force or of the nuclear wave function . in a future experiment @xcite the ratio of ( @xmath84)-pairs to ( @xmath85)-pairs knocked out from @xmath3he
will be studied with high statistics over a large @xmath34-range to further investigate the validity of the tensor - correlation dominance ( see right hand part of fig .
[ fig : srcresult ] ) .
the nucleon electro - magnetic form factors are of fundamental importance for the understanding of the nucleon s internal structure .
early @xmath5-experiments used ( quasi-)elastic scattering off the deuteron to extract the longitudinal deuteron response function . due to the smallness of @xmath5 , the use of different nucleon - nucleon potentials resulted in a 100% spread in the resulting @xmath5 . in the past decade a series of double - polarization measurements of neutron knock - out from a polarized @xmath6h or @xmath44he target have provided accurate data on @xmath5 .
the ratio of the beam - target asymmetry with the target polarization perpendicular and parallel to the momentum transfer is directly proportional to the ratio of the electric and magnetic form factors , @xmath86 where @xmath87 and @xmath88 denote the polarization component perpendicular and parallel to @xmath89 .
a similar result is obtained with an unpolarized deuteron target when one measures the polarization of the knocked - out neutron as a function of the angle over which the neutron spin is precessed with a dipole magnet .
experiment e02 - 013@xcite measured the charge form factor of the neutron in 2005 by studying the spin asymmetry in the reaction @xmath90 at four values of @xmath11 up to 3.5 gev@xmath6 .
the scattered electron and the knocked - out neutron were detected in coincidence using the open - geometry electron spectrometer bigbite and the large neutron detector bighand .
the bigbite spectrometer ( with a solid - angle acceptance of roughly 76 msr ) was equipped with a detector package containing three wire chambers and a lead - glass shower counter , separated into a pre - shower and full - absorption region .
the front and back wire chambers each had six planes of sensitive wires , the middle chamber three planes .
a hit above a high threshold in the shower counter together with a neutron event from bighand provided the basic coincidence trigger for the experiment .
the bighand neutron detector ( with an active detection area of around @xmath91 and a total weight of @xmath080 tons ) contained over 200 neutron bars arranged in seven vertical walls sandwiched between iron .
it also contained two veto walls with @xmath0180 veto counters protected from the target by two inches of lead .
the polarized @xmath44he target that sat on the pivot in hall a was basically designed from the ground up .
a magnetic holding field of roughly 2 mt was produced using a large iron box that also provided shielding from the fringe field of bigbite and served as a scattering chamber .
the sealed glass target cells in which the @xmath44he was polarized contained a mixture of potassium and rubidium , in contrast with previous target cells in which the only alkali - metal present was rubidium .
this hybrid optical pumping technique yielded a substantially higher polarization , quicker pump - up times , and less sensitivity to depolarization from the passage of the electron beam .
light from high - power diode - laser arrays was brought to the target using optical fibers .
the optics used for polarizing the light and focusing it onto the target cells were mounted directly on top of the target enclosure .
the target held a polarization of about 50% during many weeks of continuous running with an 8 @xmath1a electron beam . with the 40 cm long target cells containing @xmath44he at roughly 10 atm ,
e02 - 013 operated at a luminosity of @xmath92 /cm@xmath6/s .
with such a high luminosity , improved target polarization , and the large acceptance of the bigbite / bighand combination , a figure - of - merit was achieved that was at least 15 times larger than that of any previous @xmath93 experiment .
the analysis of the bigbite optics resulted in a vertex reconstruction of 5 mm and a momentum resolution of @xmath01% .
the shower detector was calibrated before the experiment with cosmic rays .
a resolution of @xmath00.5 ns after corrections was achieved for the time of flight ( tof ) measurement in bighand .
the veto efficiency per counter is about 95% , consistent with expectations at the high counting rate .
quasi - elastic events were selected through cuts on the invariant mass @xmath15 and on the transverse component of the missing momentum @xmath94 .
the dilution of the measured beam asymmetry due to proton leakage through charge - exchange reactions was extracted from measurements on various targets and found to be in good agreement with simulations .
a preliminary analysis of the data at q@xmath6 = 1.8 gev@xmath6 has yielded a value of @xmath5 close to the belitsky scaling prediction with an accuracy of @xmath015% , which is the total error expected for each of the four @xmath95-values , as illustrated in the left part of fig .
[ fig : expected - result ] . with the values of @xmath11 in e02 - 013 and the expected accuracy , shown relative to the galster fit .
the solid curve is a calculation in the relativistic constituent quark model by g. miller @xcite .
the dashed curve takes its shape from the pqcd prediction by a. belitsky _ et al._@xcite , normalized to the experimental @xmath93 value at 1.3 gev@xmath6 .
right : the radial distribution of the charge in the neutron , extracted from an analysis of the world data.[fig : expected - result ] ] in the breit frame the nucleon form factors can be written as fourier transforms of the charge and magnetization distributions . however , if the wavelength of the probe is larger than the compton wavelength of the nucleon , i.e. if @xmath96 , the form factors also contain dynamical effects due to relativistic boosts . the negative lobe at @xmath01 fm in the radial charge distribution shown in the right part of fig .
[ fig : expected - result ] supports the model of the neutron in which for part of the time it consists of a @xmath97 circling a proton , which is consistent with miller s calculation that shows a large contribution of the pion at low @xmath11-values .
also , at these @xmath11-values the relativistic corrections in the fourier transforms are expected to be small .
the generalized parton distributions ( gpd ) relate the spatial and momentum distributions of a parton in a nucleon by providing a consistent framework for the emff and the parton distribution functions ( pdf ) . deeply virtual compton scattering ( dvcs ) is the most accessible reaction to study gpds .
the e00 - 110 experiment@xcite ran in 2004 with 5.75 gev electrons incident on a 15 cm long liquid h@xmath98 target .
the luminosity was typically @xmath99/cm@xmath6/s with a @xmath100 beam polarization .
the scattered electrons were detected in one high resolution spectrometer ( hrs ) , while photons above a 1 gev energy threshold ( and @xmath101 coincidences from @xmath35 decay ) were detected in a @xmath102 array of @xmath103 cm@xmath44 pbf@xmath98 crystals , whose front face was located 110 cm from the target center .
dvcs events were selected from electron - photon coincidences , after subtraction of the @xmath35 yield - estimated from two - photon events - and application of a missing - mass cut @xmath105 . , and @xmath106 , as a function of @xmath107 , in the bin @xmath108 gev@xmath6 at @xmath109 .
the solid lines show total fits with one-@xmath110 statistical error bands .
the dot - dot - dashed line is the @xmath111 contribution to @xmath106 .
the short - dashed lines are the contributions from the fitted @xmath112 and @xmath113 parts of @xmath114 .
the long - dashed line is the fitted @xmath115(\mathcal f)$ ] term .
the dot - dashed curves are the fitted @xmath112 and @xmath113 parts of @xmath116.[fig : sigmadiff ] ] to order twist-3 the dvcs helicity - dependent ( @xmath117 ) and helicity - independent ( @xmath118 ) cross sections are given by @xcite : @xmath119 = \\ & = & \sin(\phi_{\gamma\gamma } ) \gamma_{1}^{\im } \
, \im { \rm m } \left[{\mathcal c^i}({\mathcal f})\right ] - \sin(2\phi_{\gamma\gamma } ) \gamma_{2}^{\im } \ , { \im { \rm m}}\left[{\mathcal c^i}({\mathcal f}^{\rm eff})\right]\ , , \nonumber \label{eq : dsigma } \\ { d^4\sigma\over d^4\phi } & \equiv & { 1\over 2 } \left[{d^4\sigma^+\over d^4\phi } + { d^4\sigma^-\over d^4\phi}\right ] = { d^4\sigma(|dvcs|^2 ) \over
dq^2 dx_{\rm bj } dt d\phi_{\gamma\gamma } } + { d^4\sigma(|bh|^2 ) \over
dq^2 dx_{\rm bj } dt d\phi_{\gamma\gamma } } \label{eq : dsigma } \\ & + & \gamma_{0,\delta}^{\re } { { \re e}}\left [ { \mathcal c^i}+\delta{\mathcal c^i}\right ] ( { \mathcal f } ) + \gamma_0^{\re } { { \re e}}\left[{\mathcal c^i}({\mathcal f})\right ] \nonumber \\ & - & \cos(\phi_{\gamma\gamma } ) \gamma_{1}^{\re } { { \re e}}\left[{\mathcal c^i}({\mathcal f})\right ] + \cos(2\phi_{\gamma\gamma } ) \gamma_{2}^{\re } { { \re e } } \left[{\mathcal c^i}({\mathcal f}^{\rm eff})\right]\ , , \nonumber\end{aligned}\ ] ] where @xmath107 denotes the azimuthal angle of the detected photon .
the @xmath120 are kinematic factors with a @xmath107 dependence that arises from the electron propagators of the bh amplitude .
the @xmath121 and @xmath122 angular harmonics depend on the interference of the bh amplitude with the set @xmath123 of twist-2 compton form factors ( cffs ) or the related set @xmath124 of effective twist-3 cffs : @xmath125({\mathcal f } ) = f_1{\mathcal h}-{t\over 4 m^2 } f_2{\mathcal e } - \xi^2 g_m \left[{\mathcal h}+{\mathcal e}\right],\ \label{eq : gpds3}\end{aligned}\ ] ] in which @xmath126 , @xmath127 and @xmath128 denote the elastic form factors .
@xmath117 measures the imaginary part of the bh - dvcs interference terms and provides direct access to gpds at @xmath129 , while @xmath118 determines the real part of the bh - dvcs interference terms and measures the integral of gpds over its full domain in @xmath8 .
-dependence of the @xmath112 parts of ( twist-2 ) @xmath130 and ( twist-3 ) @xmath131 angular harmonics , averaged over @xmath132 .
the horizontal line is the fitted average of @xmath133 $ ] .
right : extracted real and imaginary parts of the twist-2 and twist-3 angular harmonics as a function of @xmath132 .
[ fig : dvcscoef ] ] figure [ fig : sigmadiff ] shows @xmath117 and @xmath118 for one @xmath134 bin@xcite .
clearly , the twist-3 terms make only a very small contribution to the cross sections .
note also that @xmath118 is much larger than the bh contribution alone , especially from 90@xmath135 to 270@xmath135 .
this indicates that the beam spin asymmetry ( bsa = @xmath136 ) can not be simply equated to the imaginary part of the bh - dvcs interference divided by the bh cross section .
figure [ fig : dvcscoef ] ( left ) shows the @xmath11-dependence of the ( twist-2 ) angular harmonic @xmath137 $ ] over the full @xmath132 domain .
the absence of a @xmath11 dependence of @xmath133 $ ] within its 3% statistical uncertainty provides crucial support to the dominance of twist-2 in the dvcs amplitude .
@xmath133 $ ] is thus a direct measurement of the linear combination of gpds .
figure [ fig : dvcscoef ] ( right ) displays the twist-2 @xmath138 angular harmonics as a function of @xmath132 , together with the vgg model estimates @xcite .
the vgg model is in qualitative agreement with the @xmath139 $ ] data , but significantly under - predicts the @xmath113 parts of the angular harmonics . following the e00 - 110 experiment , the e03 - 106 experiment@xcite explored dvcs off the neutron .
because of the very small magnitude of the dirac form factor @xmath126 in the neutron case , this observable is supposed to be sensitive to @xmath140 , the least constrained gpd . within the impulse approximation the main contributions to electroproduction of photons on the deuteron come from coherent ( d - dvcs ) and incoherent ( p - dvcs ) and ( n - dvcs ) scattering .
the n - dvcs cross - section difference is obtained by first subtracting the proton contribution measured on a hydrogen target and then separating the neutron and deuteron contributions in each @xmath132-bin via a global analysis , that used the dynamical separation ( @xmath141 ) between these two channels and the different kinematical factors that dominate their @xmath142 moments .
the resulting neutron moments turn out to be small and negative ( fig .
[ fig : neut ] ) , consistent with theoretical expectations ; the sensivity to the gpd @xmath140 through the quark angular momentum is also shown on the figure .
a correlated constraint on the orbital angular momenta of the valence quarks , @xmath143 and @xmath144 , is then extracted@xcite from a fit with the vgg model@xcite .
-dependence of the extracted @xmath145 moments for incoherent n - dvcs , compared to vgg calculations for different values of the orbital angular momenta of the valence quarks , @xmath143 and @xmath144 ; right : experimental constraint on @xmath143 and @xmath144 from the present n - dvcs results@xcite .
a similar constraint from hermes@xcite and the result from a lqcd - based calculation@xcite are also shown.[fig : neut ] ]
mainly preliminary results were presented for eight recent jlab experiments . the bonus detector has been successfully commissioned and will provide detailed information on the neutron structure , unbiased by nuclear corrections .
polarization transfer in @xmath3he has shown to be a sensitive test of possible modifications of the nucleon form factors by the nuclear medium .
a new study of the primakoff effect has yielded accurate information on the @xmath146 decay width .
a new class of hypernuclear spectroscopy studies with much improved energy resolution has been initiated thanks to the excellent beam quality at jlab .
triple coincident nucleon knock - out studies have yielded quantitative information on short - range nucleon - nucleon correlations .
the analysis of experiment e02 - 013 will extend the existing data on @xmath5 to double the @xmath11-range .
experiment e00 - 110 has yielded the first measurements of the dvcs cross section in the valence quark region .
the @xmath11-dependence of the angular harmonics of the helicity - dependent cross section provided solid evidence of twist-2 dominance in dvcs .
the unexpectedly large contribution of the dvcs@xmath6 term observed in the cross section impedes the direct extraction of gpds from bsa measurements .
the author gratefully acknowledges detailed discussions with drs .
h. fenker , a. gasparian , j. lerose , d. higinbotham , b. wojtsekhowski and c. hyde . the permission by the spokespersons of the experiments selected to present preliminary results is greatly appreciated .
this work was supported by doe contract de - ac05 - 06or23177 , under which jefferson science associates , llc , operates the thomas jefferson national accelerator facility .
99 h. fenker , c. keppel , s. kuhn and w. melnitchouk , jlab experiment e03 - 012 l. lapikas , s. dieterich _ et al . _ ,
s. strauch _ et al .
_ , r. ent , r. ransome , s. strauch and p. ulmer , jlab experiment e03 - 104 v. punjabi _
et al . _ , o. gayou _
et al . _ , ; m.k .
et al . _ , j.m .
et al . _ , ; r. schiavilla _
et al . _ , d. dale , s. danagoulian , a. gasparian and r. miskimen , jlab experiment e02 - 103 h.w .
_ , a. browman _
_ , _ sov .
jetp _ 30 ( 1970 ) 1037 g. bellettini _
_ , _ il nuovo cimento _ 66 ( 1970 ) 243 j.l .
goity , a.m. bernstein and b.r .
holstein , b.l .
ioffe and a.g .
oganesian , o. hashimoto and h. tamura , _ prog .
_ 57 ( 2006 ) 564 o. hashimoto , s. nakamura , j. reinhold and l. tang , jlab experiment e01 - 011 t. miyoshi _
et al . _ ,
o. hashimoto , s.n .
nakamura , j. reinhold and l. tang , jlab experiment e05 - 115 s. frullani , f. garibaldi , j. lerose , p. markowitz and t. saito , jlab experiment e94 - 107 m. iodice _
et al . _ , t. mizutani _ et al . _ ,
f. cusanno _
et al . _ , in preparation for _ phys .
_ d. rohe _ et al .
_ , k. egiyan _ et al .
_ , ; o. benhar , d. day and i. sick , j.m .
laget , l.l .
et al . _ ,
et al . _ , w. bertozzi , e. piasetzky , j. watson and s. wood , jlab experiment e01 - 015 r. shneor _ et al .
_ , r. subedi _
_ , in preparation for _ science _
r. schiavilla _ et al . _ , m. m. sargsian _ et al .
_ , s. gilad , d. higinbotham , b. moffit , e. piasetzky and j. watson , jlab experiment e07 - 006 g. cates , n. liyanage and b. wojtsekhowski , jlab experiment e02 - 013 g. a. miller , a. belitsky , x. ji and f. yuan , p. bertin , c. hyde , r. ransome and f. sabati , jlab experiment e00 - 110 c. muoz camacho _
_ , a. v. belitsky , d. mueller and a. kirchner , m. vanderhaeghen , p. a. m. guichon and m. guidal , p. bertin , c. hyde , f. sabati and e. voutier , jlab experiment e03 - 106 m. mazouz _
et al . _ ,
z. ye , in m. kuze , k. nagano and t. tokushuku , editors , _ proc . of the 14th int .
workshop on dis , tsukuba , 2006 _ , world scientific , singapore , 2008 , p. 679
et al . _ , accepted for publication in , arxiv:0705.4295 [ hep - lat ] | preliminary results of a selected few recent experiments at jlab are presented .
= 11.6pt |
the launch of _ xmm - newton _ and _ chandra _ both carrying on - board diffraction grating spectrometers with high resolving power , dramatically improved our knowledge of the gaseous environment of active galactic nuclei . at first , very bright and nearby seyfert galaxies were explored and many results obtained in the last decade . depending on the observer s line - of - sight , different features could be detected . in x - ray
obscured seyfert 2s high - resolution observations have revealed soft x - ray spectra dominated by line emission from h - like and he - like elements as neon , oxygen , carbon , as well as from fe l - shell transitions ( fe xvii to fe xxiv ) ( sako et al . 2000 ; kinkhabwala et al .
2002 ; bianchi et al . 2005 ; pounds & vaughan 2006 ; guainazzi & bianchi 2007 ) . the soft x - ray emission is most likely produced by photoionisation in a gas located in the narrow line region ( nlr ) ( bianchi , guainazzi & chiaberge 2006 ) .
signatures of outflowing ionised gas have been detected in the x - ray spectra of about 50@xmath5 of nearby seyfert 1s .
the data indicate the presence of a multi - phase gas with an ionisation parameter @xmath6=@xmath7 , _
l _ is the 1 - 1000 rydberg source luminosity , _
n@xmath8 _ is the gas density and _ r _ is the distance from the source ( tarter tucker and salpeter 1969 ) ] spanning three orders of magnitude and column densities ranging between 10@xmath9 @xmath1 , with velocities of the order of a few hundreds to thousands km s@xmath3 ( blustin et al .
2005 ; mckernan et al .
2007 ) . where these winds originate and are powered is still an open question .
several regions at different distances from the central engine , like the nlr , the obscuring torus and the accretion disk have been proposed as a possible location for the gas ( crenshaw et al . 2003 and references therein ) . for radio galaxies
the investigation of the nuclear environment through high - resolution spectroscopy has just started .
there are some hints that in obscured rgs ( seen at high inclination angles ) the soft x - ray spectra are similar to their radio - quiet ( rq ) counterparts , seyfert 2s .
evidence for photoionised gas has already been revealed by emission lines in the soft x - ray spectra of 3c 445 ( grandi et al .
2007 ; sambruna et al . 2007 ) and 3c 33 ( torresi et al . 2009 ) . in the latter case ,
the morphological coincidence between soft x - ray and optical [ oiii]@xmath105007 emission suggests the nlr as the location of such gas ( torresi et al .
2009 ) .
the current picture is less clear for the radio - loud ( rl ) counterparts of seyfert 1s , where doppler amplification of the jet flux can play a key role in blurring the spectral features : for example , rgs are known to have weaker compton reflection signatures and weaker fe k@xmath11 lines when compared to seyfert 1s ( grandi , malaguti & fiocchi 2006 and references therein ) . up to now , the search for warm absorbers in unobscured rgs ( seen at low inclination angles ) has been unsuccessful , except for a claim of its existence in 3c 382 and 3c 390.3 from analysis of asca data ( reynolds 1997 ) .
therefore it is completely unclear whether winds can form and/or survive in the nuclear regions of rl agns , where large amounts of the gravitational power is funneled into relativistic collimated plasma .
addressing this point is fundamental to clarify the physical link among accretion flows , winds and jet , and the influence that each of these components can have on the host galaxy .
3c 382 is a nearby radio galaxy ( z=0.0579 ) , with lobe - dominated fanaroff
riley ii radio morphology .
it exhibits a long jet extending north east of the core for 1.68 and two radio lobes , with a total extension of 3 ( black et al .
+ at optical wavelengths 3c 382 shows strong and broad lines ( fwzi@xmath1225,000 km s@xmath3 ) , which are variable on timescales of months to years .
an _ hubble space telescope ( hst ) _
wfpc2 image showed that this source is an elliptical galaxy strongly dominated by an unresolved nucleus ( martel et al .
the optical emission of this source is clearly photoionised as can be inferred from the de - reddened line ratios , i.e. log[oiii]/h@xmath13=0.90 and log[nii]/h@xmath11=0.17 ( buttiglione et al .
2009 ) , compared with the diagnostic diagrams of miller et al .
+ in the x ray band 3c 382 is a bright source ( f@xmath14 3@xmath1510@xmath16 erg @xmath1 s@xmath3 , from this _ xmm - newton _ observation ) .
the 210 kev spectrum is well fitted with a single power - law ; when the fit is extrapolated to lower energies a strong soft - excess is observed ( prieto 2000 ; grandi et al .
rosat / hri observations revealed extended x - ray ( 0.2 - 2.4 kev ) emission around 3c 382 ( prieto 2000 ) , that was confirmed by the spatial resolved study of gliozzi et al . ( 2007 ) . however , as discussed by grandi et al .
( 2001 ) and gliozzi et al .
( 2007 ) the extended thermal emission can not account for the observed soft - excess , which must be instead of nuclear origin .
+ here we report the first high - resolution detection of an x - ray warm absorber in this blrg .
3c 382 was observed by _ xmm - newton _ on 8th april 2008 ( obsid : 0506120101 ) for about 40 ks . the data from epic ( strder et al .
2001 ) , rgs ( den herder et al . 2001 ) and optical monitor ( om , mason et al .
2001 ) instruments were operating .
as we are primarily interested in the warm absorber features , only the rgs data ( 638 @xmath17 ) are presented in this paper .
data were processed with the task _ rgsproc _ of the sas 8.0.0 , which combines the event lists from all rgs ccd , produces source and background spectra using standard extraction regions , and generates response matrices .
the background is estimated taking into account events from a region spatially offset from the source .
all spectral fits presented in this work include absorption due to a line - of - sight galactic column density of n@xmath18=7.0@xmath19 @xmath1 ( kalberla et al .
the cosmological parameters used throughout the paper are @xmath20 km s@xmath3 mpc@xmath3 , @xmath21 , @xmath22 ( spergel et al .
the spectral analysis was performed using the _ xspec _ v.11.3.2 package ( arnaud 1996 ) .
c c c c c c + transition & @xmath10(rest ) & @xmath10(lab ) & flux & ew@xmath23 & @xmath24c @xmath25 + - & ( @xmath26 ) & ( @xmath26 ) & ( 10@xmath27photons @xmath1 s@xmath3 ) & ( m@xmath26 ) & - + + nex & 12.07 & 12.13 & - & 67@xmath28 & @xmath24c=7 + + fexx & 12.78 & 12.82 & - & 43@xmath29 & @xmath24c=8 + + oviii ly@xmath11 & 18.90 & 18.97 & - & 24@xmath30 & @xmath24c=11 + + oviii ly@xmath11(em ) & 18.99 & 18.97 & 4.70@xmath31 & -23@xmath32 & @xmath24c=6 + + + + +
the first - order unbinned spectra from the two rgs cameras were simultaneously fitted in _
xspec _ using the c - statistic ( cash 1979 ) .
the quoted errors correspond to 90@xmath5 confidence level ( @xmath24c=2.71 ) for one interesting parameter .
the 6 - 35 @xmath26 ( 0.35 - 2 kev ) continuum is well fitted by a single absorbed power - law with @xmath33 ( c=5264 for 4636 d.o.f . ) .
a quite steep power - law is not unexpected in this source known to have a strong soft excess .
we investigated whether an extended thermal component ( apec ) is required by the data , as suggested by the _
chandra _ analysis .
the fit did not improve significantly , nor did the shape of the residuals ( c=5261 ) .
therefore we decided not to include any other continuum components .
however , as discussed below , we checked for the presence of weak collisional emission lines . to start with a phenomenological approach ,
the spectra were divided into a number of regions containing about the same number of bins each ( 100 ) .
we performed an accurate inspection of every single region . where the spectra showed a particularly evident absorption feature , a gaussian profile with 0-width and a negative normalization was added to the model .
only lines yielding an improvement in the fit higher than @xmath24c@xmath125 ( corresponding to @xmath12 95@xmath5 for two interesting parameters ) were taken into account .
for these structures the equivalent width was estimated and the corresponding atomic transition identified through several atomic databases such as chianti ( dere et al .
1997 ; landi et al .
2006 ) , atomdb webguide and nist .
the parameters of the most prominent lines are reported in table 1 and the phenomenological fit is shown in figure 1 .
two other absorption lines probably corresponding to fexxi / mgxi and siv / nvii transitions , although observed , are not reported in table 1 because of their marginal detection ( @xmath24c = 5 ) .
none of the detected features is due to local ( galactic ) absorption .
finally we clearly detected an emission line ( @xmath24c=6 ) identified as oviii ly@xmath11 ( see table 1 ) .
this is not an unusual result .
emission features are often observed in sources showing signatures of warm absorber ( blustin et al . 2002 ; turner et al .
2004 ) .
once the presence of absorption lines was ascertained via the phenomenological analysis , the next step was to study the physical properties of the gas using a self consistent modeling of the spectrum .
extensive simulations were carried out using the photoionisation code xstar ( v2.1ln9 ) .
the output was a grid of models ( afterwards simply called xstar model ) that can be fitted to the data in xspec .
each model in the grid represents a photoionised spectrum characterized by a particular value of the column density and the ionisation parameter .
our explored parameter range is n@xmath18=[10@xmath34 , 10@xmath35 and log@xmath6=[-3 , 3 ] .
we assumed a simple spherical geometry for the illuminated gas ( covering factor = 1 ) , solar abundances ( grevesse & sauval 1998 ) , and an illuminating continuum from 1 to 1000 rydbergs derived from the intrinsic spectral energy distribution sed of the source , with l@xmath36 erg s@xmath3 ( figure 2 ) .
the turbulent velocity of the gas was set to v@xmath37=100 km s@xmath3 , as the limited rgs statistics did not allow to constrain the intrinsic width of the absorption features .
the xstar model obtained from these simulations was applied as an _ xspec _ multiplicative component to the power - law and fitted to the data .
the fit significantly improved ( @xmath24c=160 ) , as xstar can take into account any possible absorption feature in the spectrum .
we note that the photon index is completely consistent with the one obtained from a power - law only fit .
the presence of a highly ionised gas phase is confirmed , indeed the best fit required a high ionisation parameter log@xmath6=2.69@xmath38 erg cm s@xmath3 and a column density n@xmath18=3.19@xmath39 @xmath40 @xmath1 .
the robustness of these results is attested by the n@xmath18- log@xmath6 confidence contours ( 68@xmath5 90@xmath5 99@xmath5 ) shown in figure 3 ( _ upper panel _ ) .
we attempted to estimate the gas outflow velocities by measuring the line shifts between theoretical and observed values of the energy centroid allowing the redshift parameter ( z ) of the xstar model to vary .
we found reasonable outflow speeds of the order of -1200@xmath41 km s@xmath3 , consistent with typical outflow velocities in seyfert 1s ( blustin et al .
2002 ; blustin et al .
2005 ) when the absorption gas is seen along the line - of - sight .
incidentally we point out that the phenomenogical identifications proposed in table 1 are confirmed by the xstar model ( see figure 3 _ lower panel _ ) .
we note that the energy centroid of the oviii ly@xmath11 emission line does not require any blueshift ( although with large uncertainties ) .
however it may suggest that , contrary to the absorption lines observed along the line - of - sight , the emission comes from an extended region around the source , maybe a shell , expanding in all directions .
the high - resolution analysis of the blrg 3c 382 has revealed that relativistic collimated plasma and slower outflows of photoionised gas can coexist in the same object .
the outflowing gas is found in a hot phase as attested by the derived ionisation parameter log@xmath6=2.69 erg cm s@xmath3 deduced with the xstar simulations .
most of the light elements are indeed completely ionised , and the strongest absorption features correspond to h - like elements and l - shell iron states . in seyfert
1s warm absorbers are complex and multi - phase , often indicating that layers of gas at different ionisation states are part of the same outflow ( blustin et al .
2005 ) . in the present case ,
only one component of hot gas was observed , without apparent traces of a colder component .
nonetheless , the physical conditions of the gas in 3c 382 warm absorber are quite similar to those measured in the highest ionised absorber in their radio - quiet counterparts ( ngc 4051 , krongold et al . 2007 ; mkr 279 , costantini et al . 2007 ;
iras13349 + 2438 , sako et al .
taking into account the short exposure time and the very simple assumptions in the xstar model , the result appears encouraging .
we are aware that setting v@xmath37 to an _ a priori _ fixed value could potentially overpredict the column density if the real turbulent velocity is larger .
interestingly , also the ews of the absorption lines detected in 3c 382 are consistent with those generally measured in seyfert 1s ( yaqoob et al . 2003 ; turner et al .
this similarity suggests a negligible jet contribution to the soft x - ray spectrum , although the large uncertainties in ew prevent more quantitative statements from being made . in order to localize the warm absorber we estimate a minimum and a maximum distance for the gas following blustin et al .
the minimum distance is calculated with the assumption that the outflow velocity has to be greater or equal than the escape one , @xmath42 . for a m@xmath43=1.1@xmath44 m@xmath45 ( marchesini et al .
2004 ) , and average outflow velocities of the order of 10@xmath4 km s@xmath3 ( see table 1 ) we deduce that the gas should be located at a distance larger than 10 pc . on the other hand ,
the maximum distance of the gas can be obtained by combining the ionisation parameter and the column density .
if the gas is confined in a thin layer ( @xmath46 with @xmath47 the layer depth ) , and the volume filling factor equal to 1 , it turns out that @xmath48 pc . according to ghisellini & tavecchio ( 2008 ) the blr and torus radii
are expected to scale with the square root of the disk luminosity : @xmath49@xmath50 cm and @xmath51 cm ( cleary et al . 2007 ) .
thus for 3c 382 we can estimate typical sizes for the blr and dusty torus region , i.e. [email protected] pc and [email protected] pc if l@xmath54 .
these results suggest that the bulk of the highly ionised absorption gas , detected in 3c 382 , is located beyond the torus , most likely in the nlr .
interestingly , we note that this is also the probable location of the photoionised emitting gas observed in 3c 33 ( torresi et al .
2009 ) , another powerful rg seen at higher inclination angles . +
this paper is based on observations obtained with the _ xmm - newton _
satellite , an esa funded mission with contributions by esa member states and nasa .
we wish to thank the anonymous referee for valuable comments and detailed suggestions .
we thank massimo cappi and andy pollock for very useful discussions .
et thanks for support the italian space agency ( contract asi - i/088/06/0 ) . | recent high - resolution measurements suggest that the soft x - ray spectrum of obscured radio galaxies ( rg ) exhibits signatures of photoionised gas ( e.g. 3c 445 and 3c 33 ) similar to those observed in radio - quiet obscured active galactic nuclei ( agn ) .
while signatures of warm absorbing gas covering a wide range of temperature and ionisation states have been detected in about one half of the population of nearby seyfert 1 galaxies , no traces of warm absorber gas have been reported to date in the high - resolution spectra of broad line radio galaxies ( blrg ) .
we present here the first detection of a soft x - ray warm absorber in the powerful frii blrg 3c 382 using the reflection grating spectrometer ( rgs ) on - board _ xmm - newton_. the absorption gas appears to be highly ionised , with column density of the order of 10@xmath0 @xmath1 , ionisation parameter log@xmath22 erg cm s@xmath3 and outflow velocities of the order of 10@xmath4 km s@xmath3 . the absorption lines may come from regions located outside the torus , however at distances less than 60 pc . this result may indicate that a plasma ejected at velocities near the speed of light and a photoionised gas with slower outflow velocities can coexist in the same source beyond the broad line regions .
galaxies : active x - rays : galaxies warm absorber : individual : 3c 382 |
recent cmb observations @xcite indicate that the early universe has passed through an inflationary period with hubble parameter @xmath9 .
the standard theoretical setup for inflationary models generically involve some scalar inflaton fields slowly rolling down their potential .
the slow - roll that is needed to ensure a resolution to the problems of standard big bang cosmology , and consistency with the cmb results , demands @xmath10 and @xmath11 to be smaller or of order @xmath12 @xcite . in the context of simple single scalar models with potential @xmath13 , @xmath14 and @xmath6
are a measure of flatness of the potential and , specifically , @xmath15 , where @xmath16 is the effective mass of the inflaton field .
therefore , the theoretical framework invoked for inflationary model building should also provide mechanisms to protect the potential and its flatness against quantum and/or quantum gravity corrections . in physics models
we generically associate smallness and protection of a quantity like @xmath6 with an approximate symmetry , such that when the symmetry is exact this parameter ( here the effective mass of inflaton ) is zero .
supersymmetry , for example , can be such a symmetry .
it turns out that in the presence of gravity , as in our case where we are dealing with inflation models , symmetries protecting @xmath6 are all broken , inducing an inflaton mass term of order the hubble scale @xmath17 of the background .
we hence end up with an order one @xmath6 , i.e. , the @xmath6-problem . from the above discussion
it is seen that the @xmath6-problem may appear in two ways : in a top - down approach where we invoke a theory of quantum gravity like string theory for inflationary model building . or in a bottom - up approach where we take the usual field theory setup of einstein gravity plus scalar inflaton fields , assuming that this framework is valid up to planck scale @xmath18 . in the top - down approach the @xmath6-problem appears as a classical ( not loop ) effect , usually due to the interaction of the inflationary sector with the moduli stabilization sector ,
see , e.g. , @xcite . in these models
it turns out to be easy to have small @xmath14 with controlled back - reactions on it , but @xmath6 receives order one corrections .
intuitively , in these top - down setups , the @xmath6-problem can be understood as follows . with a vacuum energy of order @xmath19
all scalars , including the inflaton , will be endowed with soft masses of order @xmath20 since the supersymmetry breaking scale is not below @xmath17 . in this work , however , we will focus on the @xmath6-problem in the bottom - up approach . in the bottom - up approach the @xmath6-problem arises from _ quantum _ loop corrections to the tree - level graviton - scalar - scalar vertex . despite being non - renormalizable
, one can still apply the ( wilsonian ) effective field theory techniques to the einstein gravity theory and consider loop corrections , e.g. see @xcite . in the presence of a scalar field @xmath21
minimally coupled to einstein gravity , as in generic inflationary models , these corrections at one loop level generically involve a @xmath22 term , a term whose presence was noted long ago @xcite .
as we will review below , such a term is quadratically divergent and in the one loop effective action appears as @xmath23 where @xmath24 is the uv cutoff of the theory and @xmath25 is an order one coefficient . assuming a planckian cutoff scale , @xmath26 , in an inflationary background where @xmath27
, this term yields a correction of order @xmath28 to the inflaton mass , causing the @xmath6-problem .
the @xmath6-problem , or the hubble scale mass term for the effective inflaton field , seems quite generic and one may put forward the idea of _ kinematically _ reducing the coefficient @xmath29 . in this letter
we explore this possibility .
one obvious possibility is to choose the cutoff @xmath24 , the scale where quantum gravity effects become important , to be one or two orders of magnitude smaller than @xmath18 @xcite . in this case
the coupling constant of gravitons will be reduced like the momentum uv cutoff and the @xmath6 problem persists .
alternatively , one may explore the idea that @xmath25 is a kinematical factor that for some reason is not of order one , while the bare cutoff @xmath24 is @xmath18 .
in fact , similar suppressions are already very well known in the context of large @xmath0 gauge theories @xcite : the nonplanar part of a given feynman diagram comes with powers of @xmath5 suppression compared to the planar part of the same diagram .
as we will show similar analysis can be repeated for the theories involving large number of scalar fields minimally coupled to gravity . in particular , if we have @xmath0 number of light fields , lighter than @xmath30 , @xmath25 turns out to have a @xmath5 suppression factor . in a sense , as if , the diagram leading to is a nonplanar diagram .
this observation is closely related to the species dressed gravity cutoff scale ideas discussed in @xcite , in light of which the @xmath31 result may be interpreted as dealing with a dressed cutoff @xmath32 while @xmath33 .
inflationary models with many scalar fields have recently got attention in view of their success in providing a natural explanation for the smallness of the inflaton self - couplings ( the issue of steepness of the potential ) and for the super - planckian excursion of the effective inflaton in the field space @xcite .
this idea is not exotic to string theory motivated inflationary settings where it is quite common to have an abundant number of fields / degrees of freedom with masses below the dressed cutoff @xmath30 , see , e.g , @xcite .
even though in some of these setups , like n - flation , the individual field excursion is greater than the dressed uv cutoff , some , like gauged m - flation @xcite or multiple m5-branes inflation @xcite , remain immune to thisbeyond - the - cutoff problem . in this work
we examine the above proposed @xmath5 resolution to the @xmath34-problem .
we assume that there is a hierarchy of scales between @xmath17 , the dressed gravity cutoff @xmath35 and @xmath18 : @xmath36 which is easily achieved by e.g. @xmath37 .
this provides a window where one can safely use the standard techniques of quantum field theory and effectively deal with a system that could be described by einstein - hilbert gravity , the inflaton sector , and other heavy remnants of the theory of quantum gravity whose masses @xmath38 falls below the new gravity cutoff , i.e. @xmath39 . the outline of this work is as follows .
we consider a system of @xmath0 light scalars minimally coupled to einstein gravity and work out basic feynman rules of the theory and compute the quadratically - divergent part of the one loop contributions to the graviton propagator and graviton - scalar - scalar vertices .
we show that one loop graviton two - point function has a linear @xmath0 parametric dependence while the graviton - scalar - scalar vertex has no @xmath0 dependence . therefore ,
if we ( re)normalize the graviton two - point function , the vertex will have a factor of @xmath40 .
this latter leads to @xmath31 ( _ cf . _ ) .
we discuss how this can resolve the @xmath6-problem in the context of many - field models like n - flation @xcite or m - flation .
consider the action of @xmath0 scalars minimally coupled to gravity @xmath41 where @xmath42 is the number of scalar fields , and summation over repeated @xmath43 indices is assumed .
one or some of these scalars play the role of inflaton(s ) , and @xmath44 could be any potential that realizes slow - roll inflation at the classical level , while the rest exhibit possible remnants of the underlying quantum gravity theory .
we assume the mass of these remnants to be below our dressed cutoff @xmath45 .
the action once quantized will receive all possible corrections compatible with the symmetries of the system , in particular an @xmath46 correction which appears at one loop level . as discussed if this term comes with an order one coupling can cause the @xmath34-problem . in what follows
we show by carrying out explicit one loop calculations involving gravitons , that this term is suppressed by factors of @xmath5 , providing a setting to resolve the @xmath6-problem in the context of multi - field models of inflation . in our analysis in this section and section 3
we ignore the loops involving scalar self - interactions .
as we will discuss in the discussion section these diagrams do not change our main result . to perform the one loop analysis , as in any quantum field theory , we need to work out basic tree level feynman diagrams of propagators and interaction vertices .
to do so for the gravity sector , following @xcite , we introduce the tensor densities , @xmath47 to bring the gravitational part of the action to the goldberg s form @xcite @xmath48 we will decompose the density metric to the flat part and the deviation from the flat space part , @xmath49 where @xmath50 is defined as @xmath51 the inverse of @xmath52 is given by @xmath53 where on the r.h.s .
the indices are raised and lowered by the flat minkowski space metric @xmath54 .
perturbing the action up to third order in @xmath55 , one obtains @xmath56 where @xmath57 note that @xmath58 is written to lowest order in @xmath55 and so it is independent of @xmath55 .
grp ( 0,0)(0,5 ) ( 40,30 ) + spr ( 0,5)(0,5 ) ( 40,30 ) + hpp0 ( 0,10)(0,20 ) ( 80,80 ) + hhpp ( 0,20)(0,20 ) ( 80,80 ) + hhh0 ( 0,20)(0,0 ) ( 80,80 ) from this interaction term , and dropping the last term in which is inessential for our purposes , one can read the vertex @xmath59 to be @xmath60 , where @xmath61 and @xmath62 are two external four - momenta on the @xmath63 particles . to work out the basic feynman graphs of the theory we need to gauge - fix the diffeomorphism invariance .
this may be done through gauge fixing term @xmath64 in figure [ hphiphi1 ] we have plotted the basic feynman graphs in this gauge .
having the tree level theory we now proceed to the one loop analysis and revisit the one loop propagator calculations as well as graviton - scalar vertex .
since we are interested in the quadratically divergent term , it is appropriate to use cutoff regularization ; dimensional regularization , which is very well suited in capturing the logarithmic divergences and already used in @xcite , can not be employed here .
hh1 ( 10,10)(30,0 ) ( 100,100 ) hh2 ( 30,0)(10,0 ) ( 100,100 ) hh3 ( 10,10)(30,10 ) ( 100,100 ) hh4 ( 30,10)(10,10 ) ( 100,100 ) t90 ( 10,0)(10,0 ) ( 80,80 ) as depicted in figure [ 1-loop - gr ] , there are five feynman diagrams contributing to the one loop graviton propagators .
the first two are coming from the pure einstein gravity sector and the other three involve scalar fields running in the loop .
since we are only interested in the @xmath0 dependence of the diagrams we focus on the ones with scalar fields in the loop . in electrodynamics , the gauge invariance enforces the photon self - energy to be transverse .
this reduces the degree of divergence from two to zero .
however , in gravity the gauge invariance does not do so .
it only relates the three diagrams that involve the scalar field in the loops .
thus the quartic divergence , @xmath65 , which corresponds to the cosmological constant term , @xmath66 , remains .
this difference is due to the fact that @xmath66 is still gauge - invariant whereas @xmath67 is not .
this is the famous cosmological constant problem which we are not intended to deal with here .
next - to - leading divergent part diverges like @xmath68 and this is the part that renormalizes the graviton propagator .
in particular , the diagram that involves two graviton - scalar three - vertices is @xmath69 as long as @xmath70 .
the diagram involving the graviton - scalar four vertex is of the form @xmath71 the leading part of this integral is quartic in the uv cutoff and just renormalizes the cosmological constant .
the next - to - leading order is quadratic in the uv cutoff and is proportional to @xmath72 , assuming that the masses @xmath38 are all much smaller than the undressed ( bare ) uv cutoff @xmath24 , which is taken to be @xmath18 .
the last diagram in figure [ 1-loop - gr ] has only a quartic divergence and does not contribute to the renormalization of the graviton propagator at all .
thus we see that the one loop graviton propagator is proportional to number of fields @xmath0 , as well as @xmath73 .
this term may be viewed as the correction to the newton constant or @xmath18 .
that is , the quantum gravity effects become important when this term becomes of the same order as the classical tree level value .
this happens if the cutoff momentum @xmath24 is of order @xmath74 which is the _ species dressed uv cutoff_. besides this `` perturbative '' argument , the fact that one should use this reduced cutoff instead of @xmath18 in presence of large number of species has also been backed up by black hole physics and hawking radiation from black holes in theories with large number of light species @xcite .
one may also compute one loop correction to the scalar propagator .
again there are diagrams involving only scalars and two diagrams involving gravitons .
it is immediate to see that the latter two diagrams have no parametric dependence on the number of scalars @xmath0 . as depicted in figure [ scalar - gravitons - exchange ] , there are three diagrams contributing to scalar - graviton vertex at one loop level .
[ c]cccccccccccccccccccccccc hpp1 ( 40,0)(40,0 ) ( 100,100 ) hpp2 ( 0,0)(0,0 ) ( 100,100 ) hpp3 ( 30,30)(40,0 ) ( 100,100 ) & & & & & & & & & & & & & & & & & & & & & & & our interest in these diagrams are twofold : _
i ) _ we read the correction to the tree level graviton - scalar three - vertex depicted in figure 1 and , _ ii ) _ compute the coefficient of the @xmath46 term which appears at one loop level from these diagrams .
the details of the loop calculations , which are straightforward , are given in the appendix , here we just quote the result .
since we are mainly interested in the @xmath0 dependence of the loop expressions we only focus on that issue here .
there are no @xmath0 dependence appearing in any of the diagrams in figure [ scalar - gravitons - exchange ] , and these diagrams , compared to the tree level results , are proportional to @xmath75 .
this in particular implies that the coefficient in front of the effective @xmath46 term ( up to numeric factors of @xmath76 ) is proportional to @xmath77 and if @xmath24 is the species dressed cutoff @xmath78 this term is suppressed by the number of light species @xmath0 . to summarize
, the one loop correction to graviton propagator is dressed with a power of @xmath0 , while the graviton - scalar - scalar vertex is @xmath0 independent .
this result is very similar to the well - established t hooft @xmath5 expansion @xcite , that if we normalize the two point function to one , the interaction term has @xmath5 suppression .
so far we have shown that in a theory with @xmath0 number of species with masses lighter than dressed cutoff @xmath78 , the coefficient of the @xmath46 term generated at one loop level is @xmath80 , where @xmath81 is an order one @xmath82-number .
the above analysis was carried out in a flat space background and should be revisited for inflationary ( almost de sitter ) backgrounds .
it is readily seen , however , that the basic argument behind the factors of @xmath0 does not depend on the background geometry .
also , the presence of the new scale @xmath83 should not affect our argument in any qualitatively important way .
we still expect that species lighter than the high energy cutoff @xmath78 in general will contribute to @xmath0 .
the only change concerns the very lightest species , with masses roughly below the hubble scale @xmath17 , where momenta at super - hubble scales will not contribute , as we will now show .
we recall that the equation of motion of a free massive scalar on an inflationary background is @xmath84 where dot denotes derivative with respect to the cosmic comoving time and @xmath85 is the scale factor .
the relevant observation is that the modes contributing to @xmath0 are the _
quantum modes _ , e.g. those with oscillatory ( as opposed to exponentially damping or growing ) behavior . to be able to solve the above equation ,
let us drop the time - dependent piece @xmath86 for the moment and consider the equation @xmath87 whose solution is of the form @xmath88 with @xmath89 . to have a quantum mode @xmath90 should have an imaginary part .
this latter implies that @xmath91 .
note that this result is @xmath92 independent and that addition of the @xmath86 term will only slightly modify this result , as it is positive definite : all the modes with @xmath93 , regardless of their @xmath92 , are always quantum modes , while modes with @xmath94 are _ classical _ for large wavelengths ( i.e. super - hubble physical momenta @xmath95 ) , and quantum mechanical for sub - hubble momenta .
note also that the damping coefficient @xmath96 is removed in the process of canonical quantization as canonical momentum conjugate to @xmath63 is @xmath97 .
this is of course the standard established result in inflationary cosmic perturbation theory and quantum field theory on curved ( de sitter ) space time @xcite .
since we are only interested in the uv behavior of the loop integrals , we can instead of integrating over @xmath92 all the way from zero to @xmath98 restrict the integral to go from @xmath17 to @xmath99 . in this way we avoid the unnecessary complication with super - hubble modes . in summary , all the modes lighter than @xmath99 , with both super - hubble or sub - hubble masses , contribute to the @xmath0 in the loop integral .
in other words , as long as @xmath100 , @xmath0 is the same for inflationary and flat space and @xmath101 where @xmath0 is the number of species lighter than the cutoff , @xmath99 . in particular
, the coefficient of the @xmath46 term generated at one loop will become @xmath80 , with @xmath81 of order one .
we are now ready to address the @xmath6-problem . to this end ,
we recall that the one loop corrected action is @xmath102 hence , the slow - roll parameter @xmath103 , where @xmath104 , is@xmath105 to have a successful slow - roll inflationary period we usually demand @xmath106 , and if we assume @xmath107 , quantum corrections to @xmath6 will be suppressed enough for @xmath108 . so that this effective inflaton field is canonically normalized . the @xmath46 term , being quadratic in the @xmath63 s , will not receive any normalization factors due to the @xmath0 scaling relating canonically normalized inflaton and the original fields . ]
the bottom - up @xmath6-problem seems quite generic to all models of inflation that involve a scalar field minimally coupled to gravity .
even if the parameters of the inflaton potential are chosen meticulously at tree - level , the loop corrections that arise from interactions of the graviton with the scalar field create the quadratically divergent conformal mass term which leads to the @xmath6-problem , if the uv cutoff of the theory is of order @xmath18 .
this kind of @xmath6-problem is of course different from the `` top - down '' @xmath6-problem arising within the string theory setups in which the volume modulus stabilization often resurrects the @xmath6-problem .
the precision that should be enforced upon the tree - level parameters are often not needed to sustain inflation , but to match the observed density of perturbation @xcite . in this letter
we examined the possibility of circumventing the @xmath6-problem , in the former sense discussed above , in many - field models of inflation that are minimally coupled to gravity .
these many - fields whose masses are assumed to be smaller than @xmath3 , have a natural appearance in effective low energy field theory description of quantum gravity models . as we argued
@xmath109 will resolve the @xmath6-problem . even though it is not necessary for our argument
, the scalar fields should be non - interacting to realize inflation .
one example of such many - field models is n - flation @xcite which has the @xmath110 symmetric potential @xmath111 .
as stated above , a few hundred scalar fields will be enough to circumvent the @xmath6-problem . like its chaotic counterpart
, the mass parameter @xmath16 has to be @xmath112 which is smaller than @xmath45 unless one resorts to an unnaturally large number of scalar fields , _
i.e. _ @xmath113 .
this of course comes at the price of exposing the model to quantum instability of the type discussed in @xcite .
namely , the quantum fluctuations of these light fields may dominate over the classical evolution of the inflaton .
assisted model with quartic potential @xmath114 is another possibility . to have an observationally viable model
, the effective coupling must be around @xmath115 and thus with @xmath116 , one needs around @xmath117 scalar fields .
this scenario also suffers from the above quantum instability with such a large number of massless scalar fields .
another disadvantage of both these two models is that the physical excursion of the fields is larger than @xmath45 during the required @xmath118 e - folds of inflation .
another explicit example is m - flation @xcite or its gauged version @xcite where the inflaton potential is constructed by three @xmath119 non - commutative hermitian matrices whose action is invariant under @xmath120 .
the classical dynamics is simplified considerably in the @xmath121 sector where these three scalar fields are proportional to the generators of the @xmath121 algebra . gauged m - flation , in addition to the above ingredients ,
has an extra @xmath122 field , associated with `` center of mass '' @xmath123 .
m - flation in the @xmath121 sector besides the inflaton field contains some number of `` spectator fields '' which do not contribute to classical inflationary trajectory while can be excited quantum mechanically and appear in the loops .
for the gauged m - flation there are @xmath124 such scalar modes and @xmath125 massive vector modes .
these modes have a hierarchical spectrum , _ i.e. _ they can be lighter or heavier than the hubble scale @xmath17 , for the explicit masses see @xcite .
not all these modes are light enough to be counted in the dressed cutoff . as discussed in @xcite , the number of `` contributing species '' @xmath126 varies between @xmath127 and @xmath128 , depending on the region of the potential inflation happens .
hence , the species dressed uv cutoff is @xmath129 .
consequently the @xmath22 term is suppressed by a factor of @xmath130 and could be safely ignored .
gauged m - flation could be motivated from the branes dynamics in an appropriate flux where the above matrices correspond to three of the perpendicular directions of a stack of @xmath0 d3-branes which are scalars in the adjoint representation of the @xmath120 . as such , although the quantum " @xmath6-problem is resolved for m - flation , embedded within string theory , one should still deal with the `` stringy @xmath6-problem '' ( cf . introduction for further references and discussions ) . finally , the term leading to the @xmath6-problem , @xmath4 , is a one - loop but marginal operator .
one may naturally worry about other loop corrections , that enhancement factors of @xmath0 will dominate over @xmath131 suppressions .
it is straightforward to show that the largest such @xmath0 enhancement factor appears in the graviton two point function ( at higher loops ) for which this factor is @xmath132 , where @xmath133 is number of loops .
all the other diagrams will be of the form @xmath134 with @xmath135 .
hence our one loop result seems to be also valid to all orders .
a.a . was supported by the gran gustafsson foundation .
. is supported by the swedish research council ( vr ) and the gran gustafsson foundation .
we would like to thank liam mcallister for helpful discussions .
now let us focus on all the diagrams which may generate the @xmath22 term , or equivalently @xmath136 , at the one - loop order .
the first one , is given by the left diagram in figure [ hphiphi1 ] , where a graviton exchange between two external scalars modifies the scalar form factor .
the diagram is proportional to @xmath137 \frac{i(\eta^{\alpha\gamma}\eta^{\beta\kappa}+\eta^{\alpha\kappa}\eta^{\beta\gamma}-\eta^{\alpha\beta}\eta^{\gamma\kappa})}{(p - k)^{2}+i\epsilon } \right .
\nonumber\\ & \left .
\frac{i}{k^{2}+i\epsilon } \frac{1}{m_{\mathrm{pl } } } \left [ p^{\prime}_{\gamma}(k+q)_{\kappa}-\eta_{\gamma\kappa}p^{\prime}\cdot ( k+q)\right ] \frac{1}{(k+q)^{2}+i\epsilon}\right .
\nonumber\\ & \left .
\frac{1}{m_{\mathrm{pl } } } \left [ k_{\mu}(k+q)_{\nu}-\eta_{\mu\nu } k\cdot(k+q)\right ] \right ) .\end{aligned}\ ] ] what we are interested in is the divergent part of the above integral which multiplies the generated @xmath22 term .
there are six momenta in the numerator , only four of which are internal
. thus the integral will be divergent as @xmath138 the other diagram that contributes to the @xmath139 vertex is the middle one in figure [ scalar - gravitons - exchange ] . to estimate the leading divergent part of this diagram
, one should note that the vertex that involves three @xmath140 is proportional to @xmath141 .
more specifically , @xmath142 \ , .\nonumber\end{aligned}\ ] ] sum is over symmetrization on the index pairs @xmath143 , @xmath144 and @xmath145 and also the six permutation done over momentum index triplets @xmath146 , @xmath147 and @xmath148 .
again , this diagram has six momenta in the numerator , two of which are external .
thus the divergent part of diagram behaves as times the terms among which @xmath22 exists .
finally , let us look at the right one - loop diagram in fig .
( [ scalar - gravitons - exchange ] ) .
the vertex @xmath149 is proportional to @xmath150 . in more details : @xmath151 \,.\nonumber\end{aligned}\ ] ] the diagram contains two propagators in the denominator and four momenta in the numerator , two of which are external . thus the leading correction will be , again , of order times the terms among which @xmath22 exists .
note that besides the conformal mass term , there are also other higher dimensional operators , whose explicit coefficients could be obtained by exactly calculating the amplitudes .
for example , from usual tensorial analysis , terms like @xmath152 are expected to be generated from such one loop diagrams . however , all these terms are suppressed by extra powers of the cutoff and also by slow - roll parameters in an inflationary background .
the other point which is worth mentioning is that the existence of higher order self - interactions for the scalar fields will not disturb our argument .
for example inclusion of mass term in the potential for the inflaton , will modify the scalar field propagator and also add corrections proportional to @xmath153 in the @xmath154 vertex .
such terms would at most introduce correction of order @xmath155 to the leading contribution .
thus , as long as @xmath156 , the effect of such terms are very small
. other forms of potential for the scalar field induce vertices that lead to more loops whose effect is more suppressed in comparison with the one - loop diagrams .
b. s. dewitt , `` quantum theory of gravity .
1 . the canonical theory , '' phys . rev . * 160 * , 1113 - 1148 ( 1967 ) ; `` quantum theory of gravity .
2 . the manifestly covariant theory , '' phys . rev . * 162 * , 1195 - 1239 ( 1967 ) .
a. ashoorioon , h. firouzjahi , m. m. sheikh - jabbari , `` m - flation : inflation from matrix valued scalar fields , '' jcap * 0906 * , 018 ( 2009 ) .
[ arxiv:0903.1481 [ hep - th ] ] ; `` matrix inflation and the landscape of its potential , '' jcap * 1005 * , 002 ( 2010 ) , [ arxiv:0911.4284 [ hep - th ] ] . | we observe that the dominant one loop contribution to the graviton propagator in the theory of @xmath0 ( @xmath1 ) light scalar fields @xmath2 ( with masses smaller than @xmath3 ) minimally coupled to einstein gravity is proportional to @xmath0 while that of graviton - scalar - scalar interaction vertex is @xmath0 independent .
we use this to argue that the coefficient of the @xmath4 term appearing at one loop level is @xmath5 suppressed .
this observation provides a resolution to the @xmath6-problem , that the slow - roll parameter @xmath6 receives order one quantum loop corrections for inflationary models built within the framework of scalar fields minimally coupled to einstein gravity , for models involving large number of fields .
as particular examples , we employ this to argue in favor of the absence of @xmath6-problem in m - flation and n - flation scenarios .
+ @xmath7 _ institutionen fr fysik och astronomi uppsala universitet , + box 803 , se-751 08 uppsala , sweden _ + @xmath8 _ school of physics , institute for research in fundamental sciences ( ipm ) , + p.o.box 19395 - 5531 , tehran , iran _ + |
the evolution of multi - particle systems arising in very diverse domains ranging from biology to material science is often governed by activated processes that occur with very low probability , but have a dramatic effect on the structure .
examples of physical phenomena controlled by such rare events are protein folding in biology , defect diffusion and crystal nucleation in condensed matter physics and cluster rearrangement in chemistry .
activated events are characterized by the fact that when they occur , their duration is rather short , e.g. of the order of the picoseconds in dense molecular systems , but that , in contrast , the typical time needed for them to start is very large .
this separation of time - scale is the very definition of metastability ; its origin may be energetic ( high barriers ) or , more likely in high dimensional systems , at least partly entropic ( pathways hard to find ) . even though physical times of activated events can be achieved in computer simulation
using molecular dynamics , monitoring rare fluctuations is computationally expensive and still remains a challenge because of the long waiting time for the event to occur . to overcome this problem ,
several different strategies have been developed over the past years .
a common idea behind these techniques is to bias the dynamics of the system in order to enhance the occurrence of reaction , or to use previous knowledge of the outcome of the reaction and to fix the endpoint of the trajectory . in order to implement this , in many cases one needs to know an order parameter able to discriminate between the reactant basin , the saddle regions and the product basin .
a set of methods involving the direct sampling of path ensembles has been developed during the last decade .
@xcite the strategy is to restrict paths to the subset of reactive paths , those that interpolate between reactant and product basins .
examples of such methods are transition path sampling , @xcite transition interface sampling .
@xcite another family of methods such as metadynamics @xcite and forward flux sampling @xcite simulate the evolution in time of the system , and include some form of bias that guarantees that the reaction happens , but then of course the effect of the bias has to be discounted in order to retrieve the true statistics .
methods differ also in the way rate constants are estimated , @xcite and the ability to handle nonequilibrium steady states .
observation of rare events and exploration of energy landscapes of a multi - particle system are indeed two interconnected problems .
the requirement of knowing the target of the reaction , and in particular a relevant order parameter that parametrizes the path is problematic in physical situations where nothing _ a priori _ is known .
thus , a preliminary step prior to the calculation of finite - temperature rate constants using the sampling techniques mentioned above may consist in locating the saddle points of the energy and the corresponding energy minima using one of the eigenvector following methods described in the literature ( e.g. the activation - relaxation technique , @xcite optim @xcite or the dimer method @xcite ) .
the technique in all these methods relies on monitoring the eigenvalues of the hessian matrix ( the matrix of the second derivatives of the potential energy ) , in order to individuate stable or unstable directions , i.e. minima or saddles .
a limitation of these methods is that energy saddles only correspond directly to the actual barriers for the dynamics if the temperature is very low , otherwise the entropic contribution to the dynamics becomes relevant .
another approach that may be used to explore the _ free - energy _
( i.e. finite - temperature ) barriers between basins @xcite is inspired by supersymmetric quantum mechanics , which has been used in the context of field theory to derive and generalize morse theory , precisely the analysis of the saddle points of a function .
transposing to statistical physics this formalism , one obtains a family of generalizations of langevin dynamics , converging to barriers and reaction paths of different kinds , rather than to the equilibrium basins .
the resulting method involves the evolution in phase space of a population of independent trajectories that are replicated or eliminated according to the value of their lyapunov exponent.@xcite the theory guarantees that by selecting trajectories having larger lyapunov exponents in a specified way , the bias is just what is needed so that the population describes the evolution of the _ transition current _ , rather than that of the configurations,@xcite as they would in an unbiased case .
the advantage is that the convergence of the current distribution is much faster than the typical passage time .
in this paper we shall present a more direct and elementary derivation of this dynamics without resorting to quantum theory ( section [ theorie ] ) . extending the more concise demonstration of ref .
, we will show how the modified dynamics precisely reproduces the evolution in the phase space of the probability current ( or more precisely , the ` transition ' probability current ) between equilibrium basins , thus achieving a _ probability current sampling _ of the system dynamics .
as the transition current evolves in time , it explores the different barriers , indicating which states are reached after different passage times . starting from an initial equilibrium configuration , far from equilibrium phenomena
are easily sampled , as the simulated current is an intrinsically out of equilibrium quantity.@xcite in section [ method ] , we discuss the actual population dynamic algorithm that may be used to simulate the probability current .
finally , in order to assess the performances of this probability current sampling , we present in section [ sec : lj38 ] an application to the structural transitions of a lennard - jones cluster , sampled under different physical conditions . we shall recover known results , and discuss some new ones , obtained in all cases at rather small computational costs .
let us consider a 3d many - body system composed of _ n _ particles in contact with a thermal bath .
we denote @xmath1 the configurational degrees of freedom and @xmath2 the interaction potential . in the limit of large friction @xmath3 , where inertia can be neglected
, the system evolves with a standard overdamped langevin dynamics @xmath4 where @xmath5 are independent gaussian white noises of unit variance and @xmath6 correspond to the particle masses . the probability to find the system at position @xmath7 then evolves with the fokker - planck equation @xcite @xmath8 where we have introduced the fokker - planck operator @xmath9 .
we can use the probability current @xmath10 to write eq . [
1.1 ] as a continuity equation for the probability density : @xmath11 for systems with separation of time scales , the dynamics can be split into two regimes .
starting from an arbitrary probability distribution , @xmath12 relaxes _ rapidly _ into a sum of contributions centered on the metastable states . at _ much longer times _ , the rare transitions between the metastable states make @xmath13 relax to the equilibrium distribution .
two time scales can also be identified for the dynamics of the probability current . while the probability density rapidly relaxes into the metastable states , the probability current converges _ on the same time scale _ to the most probable transition paths between the metastable states .
then , the late time relaxation towards equilibrium corresponds to a progressive vanishing of the current , when forward and backward flux between each metastable state balance .
@xcite note that the same line of reasoning holds for non - equilibrium systems in which the forces do not derive from a global potential . in such systems ,
the probability current never vanishes and converges instead to its steady - state value .
if one were able to simulate the evolution of the probability current , one would thus have all the knowledge relevant for the transitions between metastable states , while only having to simulate the system for relatively short time - scales ( similar to the equilibration time within a metastable state ) . as mentioned in the introduction , simulating directly the transition current is the goal of this paper and we now derive a self - consistent evolution equation for @xmath14 .
let us define the _ current operator _
@xmath15 : @xmath16 so that the probability current and the fokker - planck operator becomes @xmath17 the evolution of the probability current is then given by @xmath18 where we have assumed that @xmath9 does not depend explicitly on time .
straightforward algebra shows that @xmath19 and @xmath20 which turns equation into @xmath21 using the expressions and for the currents and fokker - planck operator we obtain @xmath22 note that the equations and are not self - contained : the knowledge of @xmath23 is required to compute @xmath14 . on the contrary ,
depends exclusively on the current , and can readily be used to simulate @xmath14 , without having to compute @xmath23 beforehand .
the only condition is that the current distribution at the initial time @xmath24 indeed derives from a probability distribution , i.e. is of the form : @xmath25 \label{initial}\ ] ] this means that the initial current distribution should be such that the quantity @xmath26 @xmath27 is a gradient , @xmath28 .
a particularly simple initial condition is obtained if one assumes that @xmath29 is gibbsean @xmath30 in a region @xmath31 , and zero elsewhere .
then , from eq .
( [ initial ] ) , @xmath32 is zero everywhere except on the surface of @xmath31 , where it takes the form of a vector normal to the surface of @xmath31 , and with amplitude proportional to the gibbs weight .
the evolution of current distribution given by equation , starting from an appropriate initial current @xmath33 converges to the stationary distribution of currents between metastable states _ on the same time scale _ as the usual langevin equation converges to metastable - state .
it is thus not necessary to wait for rare events to identify the transition path between the metastable states , an important improvement over standard md methods .
if there are several metastable states and transitions with different rates , the current distribution at longer times concentrates on the paths between regions that have not yet mutually equilibrated , and vanishes in transitions between states that have had the time to mutually equilibrate . in many physical situations inertia plays an important role and one can not rely on overdamped langevin equations .
@xcite in such case , the stochastic dynamics of @xmath34 interacting particles is given by @xmath35 [ langund ] where @xmath36 correspond to the @xmath37 velocities . in order to transpose the derivation of the current evolution given in subsection [ sec : fp ] to the inertial case , we have to consider the kramers evolution equation @xmath38 instead of the fokker - planck equation : @xcite @xmath39p\;.\ ] ] as in the overdamped case , this can be written as a conservation equation where the probability current _ in phase space _ is given by [ test ] @xmath40 once again , the current contains all the information about transitions between metastable states .
there is however a conceptual difference : the presence of inertia makes it inherently difficult ( and indeed , useless ) , to compute ( [ test ] ) as it stands .
the reason is that the _ phase - space _
current @xmath41 is non - zero even in canonical equilibrium .
for example , in a harmonic oscillator @xmath42 $ ] , the phase - space current in equilibrium turns clockwise in circles around the origin .
for this reason , the part of the current that corresponds to transitions between metastable states is screened by the large contributions of the currents within metastable states ( see figure [ fig : exprobcur ] ) .
the probability current and does not really represent the transition paths between metastable states .
we can however define a _
transition current _
@xcite [ eqn : transcur ] @xmath43 which has two interesting properties .
firstly , this current differs from the probability current by a divergenceless term and thus also satisfies the continuity equation @xmath44 .
fluxes out of a closed surface surrounding a metastable state are then the same for the probability and transition currents .
the latter current thus contains the relevant information about , for instance , transition rates .
secondly , the transition current vanishes in equilibrium , as can be checked by comparing and .
this current thus contains only the information relevant for the transitions between metastable states ( see figure [ fig : exprobcur ] ) and is _ not _ screened by the large ` equilibrium ' currents within them . in phase space
the transition path is sampled uniformly .
equilibrium contributions are not present .
, title="fig : " ] in phase space .
the transition path is sampled uniformly .
equilibrium contributions are not present . ,
title="fig : " ] in phase space .
the transition path is sampled uniformly .
equilibrium contributions are not present . , title="fig : " ] using algebra similar to that of the overdamped case
, one can show @xcite ( see the appendix [ app : derivation ] ) that the reduced current evolves with@xmath45 where the @xmath46 matrix @xmath47 is given by @xmath48 again , is a self - consistent equation for the transition current and we shall now show how it can be simulated .
the probability density @xmath49 is a scalar field that can be obtained by simulating many copies of the system , evolving their positions and velocities with the langevin equation , and constructing histograms . on the contrary
, @xmath50 is a vector field and thus can not be obtained in the same way : it also requires vectorial degrees of freedom .
we first present a population dynamics that can be used to construct the evolution of the transition current in section [ sec : evolcurtheory ] and then give the corresponding diffusion monte carlo algorithm in section [ sec : evolcuralgo ] . for systems with many degrees of freedom ,
the direct resolution of the partial differential equation yielding the evolution of @xmath50 is not achievable numerically . in the same way
as the langevin dynamics represents an alternative to the resolution of the kramers equation , we can use a stochastic dynamics that generates the current evolution numerically .
the explicit construction of the dynamics was already presented in detail in a previous paper .
@xcite the idea is to proceed in two steps .
firstly , a langevin dynamics is coupled to a carefully chosen evolution ( see below ) of a @xmath51-dimensional vector @xmath52 in order to account for the vectorial degrees of freedom of @xmath53 .
secondly , the transition current @xmath50 is given by the following average : @xmath54 where @xmath55 is a joint distribution obtained from the langevin dynamics , for instance by simulating a large number of copies of the system .
@xcite in practice , the coupling between the vectorial and phase space degrees of freedom is obtained by the following population dynamics .
one consider @xmath56 copies of the system ( called ` clones ' ) , identified by positions and velocities @xmath57 and @xmath58 , which all carry a @xmath51 dimensional unitary vector @xmath59 .
the dynamics of each clone is then as follows : @xcite * @xmath57 and @xmath60 evolve with the standard langevin dynamics with inertia * the vector @xmath59 evolves with @xmath61 * each clone has a _ birth - death rate _ @xmath62 .
this is the only way the vector @xmath59 influences the dynamics .
the distribution of clones @xmath63 then evolves with @xcite @xmath64 f - { \bf u}^\dagger { \bf m } { \bf u } f\ ] ] the first term of the r.h.s .
comes from the langevin dynamics , the second one from the evolution of the vector and the last one from the birth - death events .
the transition current is then given by .
on can indeed check that taking the time derivative of the r.h.s of and using , one recovers the evolution of the transition current ( see previous papers @xcite for more details ) .
we first present the implementation of the population dynamics of the clones and then discuss the construction of the transition current .
the @xmath65-th clone is denoted by its phase - space coordinates and vectors at time @xmath66 , @xmath67 .
we start with @xmath68 clones whose positions and vectors are arbitrarily chosen . at every time step , the dynamics is as follows : 1 .
all the vectors @xmath69 are rescaled to have a unitary norm .
the positions and velocities of the clones are propagated using a leap - frog discretized langevin dynamics .
@xcite 3 .
the vectors @xmath69 evolve with the ( leap - frog discretized version of ) the following dynamics : @xmath70 note that at after this step , the vectors are no more of unitary norm .
4 . for each clone one records @xmath71 .
we associate to each clone @xmath65 a probability weight @xmath72 and a random number @xmath73 chosen uniformly in @xmath74 . the clone is then replaced by @xmath75 copies , with @xmath76 where @xmath77 is the integer part of @xmath78 : if @xmath79 , @xmath80 new copies of the m - th clone are made . if @xmath81 , the clone is deleted and if @xmath82 , nothing happens .
the population size is thus increased by @xmath80 if @xmath79 or decreased by @xmath83 if @xmath81 .
after the step 5 , the population is rescaled from its current size @xmath84 to its initial size @xmath68 , by uniformly pruning / replicating the clones .
steps 1 to 3 correspond to the propagation step of independent clones , whereas steps 4 to 6 correspond to selection steps . in particular , the steps 4 and 5 correctly represents the cloning rate @xmath85 of the previous subsection since @xmath86 , so that @xmath87 .
the rescaling of the population at the step 6 can be done in many ways .
for instance , one can pick a new clone at random @xmath68 times among the @xmath88 clones obtained at the end of the step 5 .
we used an alternative approach that is less costly in terms of data manipulations : if @xmath89 , we kill @xmath90 clones chosen uniformly at random among the @xmath88 obtained at the end of the step 5 .
conversely , if @xmath91 , we choose uniformly at random @xmath92 clones and duplicate them . note that even though the clones evolve independently during the propagation step , their dynamics are correlated because the deleted clones are replaced by the duplicated ones at the selection steps . when the probability weights of the clones are all equal
, we have @xmath93 and @xmath82 . as a result
, the population is left unchanged .
conversely , when the clone weights take distinct values , clones with small weights are likely to be replaced by those with large weights .
there are many ways of implementing the resampling of the population ( steps 5 - 6 ) , well documented in the literature on diffusion monte carlo .
@xcite in particular , it could be advantageous to do the resampling only every @xmath94 time steps , where @xmath94 is tuned to ensure ergodicity in phase space , i.e. to achieve enhanced sampling towards the unstable regions where saddles are located .
since the clones move in phase space with a langevin dynamics , it can be surprising that they converge rapidly to the reaction paths , i.e. that they explore efficiently the transition states .
this can however be understood by noting that their dynamics ( without taking the averages ) is the so - called _ lyapunov weighted dynamics _
@xcite which is used to bias the langevin dynamics in favor of chaotic trajectories .
the clones will then tend to ` reproduce ' favorably in the neighborhood of saddles , which are particularly chaotic regions of phase space , and to die in wells .
this generates an ` evolutionary pressure ' that helps the clone escape from metastable states and find the reaction paths . as mentioned before , this dynamics
does not provide _ directly _ the transition current and one still has to construct the averages .
this can be difficult and clever methods to do so were discussed in the literature , for instance by mossa and clementi who studied the folding of chain of aminoacids .
@xcite the difficulty is connected to the well - known sign problem : large population of clones with arrows pointing in opposite directions cancel in the average but can numerically screen smaller asymmetric distribution that contains the information relevant for the transition current .
one can however show that if one starts from a population of clones uniformly spread over a reaction path separating two metastable states and pointing in the same direction , the time taken for the sign problem to occur is of the order of the tunneling time through the barrier ( see appendix [ app : ex1d ] ) . in the following
we will always simulate much shorter times and omit the averaging steps to simply look at the distribution of clones .
this distribution often suffices to locate the reaction paths , as can be seen on figure [ fig : exprobcur ] for the double well potential . to extract further information , for instance regarding the reaction rates
, we would need to do the averages , as was done in ref . , but this is beyond the scope of this article .
we now turn to the study of transitions between metastable states in the 38-atom lennard - jones cluster , a benchmark model system that has been extensively investigated in the past .
@xcite this system has a complex potential energy landscape organized around two main basins : a deep and narrow funnel contains the global energy minimum , a face - centered - cubic truncated octahedron configuration ( fcc ) , while a separate , wider , funnel leads to a large number of incomplete mackay icosahedral structures ( ico ) of slightly higher energies .
although the lowest potential energy minimum corresponds to the fcc structure , the greater configurational entropy associated to the large number of local minima in the icosahedral funnel make this second configuration much more stable at higher temperatures . as temperature increases ,
this system thus undergoes the finite - size counterpart of several phase transitions .
first , a solid - solid transition occurs at @xmath95 when the octahedral fcc structure gives place to the icosahedral ones . at a slightly higher temperature , @xmath96 , the outer layer of the cluster melts , while the core remains of icosahedral structure .
@xcite this ` liquid - like ' structure , also referred to as anti - mackay in the literature , then completely melts around @xmath97 .
@xcite the numerical study of this system is challenging : global optimization algorithms have failed to find its global energy minimum for a long time @xcite and direct monte carlo sampling fails to equilibrate the two funnels .
the study of the equilibrium thermodynamics of this system required more elaborate algorithms such as parallel tempering , @xcite basin - sampling techniques , @xcite wang - landau approaches @xcite or path - sampling methods .
@xcite more recently , the dynamical transitions between the two basins has been studied following various approaches .
the interconversion rates between the fcc and ico structures have been computed using discrete path sampling .
@xcite this elaborate algorithm relies on the localization of minima and saddles of the potential energy landscape , using eigenvector following , and then on graph transformation @xcite to compute the overall transition rate between two regions of phase space . to the best of our knowledge , this is the most successful approach as far as computing reaction rates in lj@xmath0 is concerned .
@xcite however , the numerical methods involved are quite elaborate , require considerable expertise and have a number of drawbacks , all deriving from the fact that it is based on the harmonic superposition approximation and the theory of thermally activated processes .
it thus requires any intermediate minima between the two basins to be equilibrated and this is only possible for small enough systems at low temperatures .
@xcite more importantly , when the difficulty in going from one basin to the other is due to entropic problems , as is the case for instance in hourglass shaped billiards , then the knowledge of minima and saddles of the potential energy landscape is not sufficient .
another attempt to study the transitions between the two funnels of lj@xmath0 relies on the use of transition path sampling .
@xcite because of the number of metastable states separating the two main basins , the traditional shooting and shifting algorithm failed here , despite previous success for smaller lj clusters .
@xcite the authors thus developed a two - ended approach which manages to successfully locate reaction paths between the two basins : they started from a straight trial trajectory linking the two minima , and obtained convergence towards trajectories of energies similar to those obtained in the discrete path sampling approach .
@xcite although the authors point out the lack of ergodicity in the sampling within their approach and the sensitivity on the ` discretization ' of the trajectories , this is nevertheless a progress and the main drawback remains the high computational cost ( the work needed @xmath98 hours of cpu time ) to obtain such converged trajectories .
in contrast , the simulations we present below required less than @xmath99 hours of cpu time . before presenting our simulations results , we give some technical details on the lj@xmath0 system and on the visualization of the different metastable states .
the lennard - jones potential is given by the expression @xmath100\ ] ] where @xmath101 is the position of the @xmath102-th atom , @xmath103 is the distance between atoms @xmath102 and @xmath104 , @xmath105 is the pair well depth and @xmath106 is the equilibrium pair separation . in addition , all the particles are confined by a trapping potential that prevents evaporation of the clusters at finite temperature ( i.e. particles going to infinity ) . if the distance between the position @xmath57 of a particle and the center of the trap @xmath107 exceeds @xmath108 , then the particle feels a potential @xmath109 .
lj reduced units of length , energy and mass ( @xmath110 ) will be used in the sequel so that the time unit @xmath111 is set to @xmath83 . rather than listing the 228 degrees of freedom of the atomic cluster
, configurations are traditionally described using the @xmath112 bond - orientational order parameters @xcite that allow to differentiate between various crystalline orders @xmath113 where the @xmath114 are spherical harmonics and @xmath115 are the polar and azimuthal angles of a vector pointing from the cluster center of mass to the center of the _ ( j , k ) _ bond which connects one of the @xmath116 pairs of atoms .
@xcite the parameter @xmath117 is often used to distinguish between the icosahedral and cubic structures , for which it has values around 0.02 and 0.18 respectively .
@xcite @xmath118 however does not distinguish between the icosahedral and the liquid - like phase and one thus often uses @xmath119 , for which fcc , icosahedral and the liquid - like phase take values around @xmath120 , @xmath121 and @xmath122 , @xcite respectively .
to show the spread of the various basins in the @xmath123 plane , we ran several molecular dynamic simulations , long enough to equilibrate within each basin but short enough so that one does not see tunneling ( see figure [ fig : bassinsmd ] ) . .
the positions of the phases barely move in the @xmath123 plane when the temperature changes , although their spreading does .
the kinetic energy however shifts when the temperature changes , and is roughly proportional to @xmath124 where @xmath34 is the number of degrees of freedom . ]
although the whole temperature scale is interesting , the challenging part from a computational point of view is the low - temperature regime where ergodicity is difficult to achieve .
below , we show the results of our algorithm for three temperatures : @xmath125 , @xmath126 and @xmath127 that span the ranges around the solid - solid and partial melting transitions .
given the high dimensionality of the system , it is difficult to follow the evolution of all the coordinates of the clones in order to know if they have localized interesting structures .
instead , we proceed as follows : we plot the evolution of the average over the clone population of @xmath118 , @xmath119 and @xmath128 as a function of the simulation time and we frequently store the positions and velocities of all the clones .
if we see a plateau in @xmath129 , @xmath130 and @xmath131 , two cases are possible : either the clones have converged to a reaction path , or they are stuck in a metastable basin . in order to distinguish the two situations , we run an auxiliary short molecular dynamic simulation ( _ without cloning _ ) starting from the positions and velocities of the clones .
the duration of this auxiliary simulation is long enough to observe relaxation into the metastable basins , but much shorter than the transition times .
if the clones evolve away from the region they had populated in phase - space , we know they had found a reaction path and the auxiliary md simulation converges to the metastable basins connected by this reaction path . if on the other hand the clones do not evolve away , we know that they had been stuck in some local basin
. in such a case we can change two parameters to enhance the sampling of the phase space : the number of clones and the friction @xmath132 ( see below for more details ) .
the time step is always @xmath133 .
note that this procedure could be automated , but the way to do so is let for future work . in principle ,
any observable that can measure whether the population of clones splits in two separate sub - populations after a short langevin dynamics would be suitable . if the clone population splits in two subpopulations with the same @xmath118 , @xmath119 and @xmath128 , we may fail to detect the corresponding barrier .
however , this coincidence would be extremely unlikely .
last , in addition to help us localize barriers , these short langevin simulations allows us to explore the true dynamics close to a particular transition states .
we first ran several simulations at @xmath134 , where the most stable state is the mackay icosahedral minimum ( icom ) while the liquid - like phase ( icoam ) and the fcc basin are metastable .
starting from the icom basin with @xmath135 clones and a low friction @xmath136 , the clones rapidly find ( @xmath137 ) a transition path to the liquid - like phase icoam .
later on ( @xmath138 ) an activated event bring the clones to another reaction path that points towards the fcc funnel .
these times have to be compared with the transition time between the icom and fcc basins that was previously evaluated in the literature at roughly @xmath139 .
@xcite note that each barrier or path act as a metastable state for the cloned dynamics and it is by activation that the population jumps from one barrier to another .
running the same dynamics with a larger number of clones ( @xmath140 ) tends to stabilize the first barrier so that one has to wait longer to see the transition to the second one .
starting from the fcc minimum with the same number of clones and at the same friction results in the clones rapidly going out of the fcc funnel and falling in the amorphous zone at the entrance of the icosahedral funnel .
@xcite a reaction path is _ followed _ by the clones but not _ maintained_. to stabilize this reaction path , we increased the number of clones and the friction .
the effect of the former is mostly to slow down the dynamics while the latter allow the clones to populate the reaction path more uniformly . for @xmath140 clones and @xmath141 , the population of clones indeed stabilizes the reaction path leading from the fcc basins to the entrance of the icosahedral funnel .
the reason why we need more clones to stabilize this barrier than the ones in the icosahedral funnel is probably that the former is more flat and spread than the latter ones @xcite and thus requires a larger number of clones to be sampled uniformly .
all these results are plotted on figure [ fig : clonet015 ] , in which we show the three basins icom , fcc and icoam obtained from the initial md simulations ( see figure [ fig : bassinsmd ] ) and the positions of the clones in the @xmath123 and @xmath142 plans at different simulation times .
and @xmath142 plans at @xmath134 .
the clones starting from the icosahedral basin first find the barrier between icom and icoam ( black symbols , @xmath143 ) .
they then fall back in the ico basin before finding a path that points towards the fcc funnel ( blue symbols , @xmath144 ) .
starting from fcc , the @xmath145 clones find a path that leads toward the icosahedral funnel ( green symbols , @xmath146 ) . *
right : * we plot @xmath119 as a function of time for the clones starting from the icom basins ( red symbols ) and the fcc basin ( magenta symbols ) .
arrows indicates the time at which the snapshots shown in the left and center panel are taken . ]
to identify the various metastable basins connected by the clones , we ran several short md simulations starting from the two long - lived plateaux ( blue and green arrows in the right panel of figure [ fig : clonet015 ] ) .
histograms made from these md runs are shown in figure [ fig : t015histoclones ] .
they show that the clones going out of the ico basin find barriers toward the amorphous region at the entrance of the ico funnel while the ones starting from the fcc minimum find a reaction path between the fcc basin and the icosahedral funnel .
interestingly , this path goes through a faulty fcc basin located around @xmath147 that has been previously reported in the literature .
@xcite started from the clones positions at the times indicated by the green and blue arrows in figure [ fig : clonet015 ] .
the gray - dotted regions correspond to the equilibrium md simulations of the three basins icom , icoam and fcc .
* left * : md simulations started from the stationary structures found by the clones in the ico funnel fall either back into the ico basin or in a metastable basin around @xmath148 that corresponds to an amorphous structure at the entrance of the ico funnel .
* center * : the clones starting from the stable structure found in the fcc funnel fall either back in the fcc basin , or in a faulty fcc metastable state ( blue rectangle ) or in the ico funnel , which .
both structures thus correspond to reaction paths . *
right * : position of the surface atoms of a clone that has fallen in the faulty fcc configuration after the short md .
this figure was made using a mathematica spreadsheet that can be downloaded at http://www-wales.ch.cam.ac.uk/~wales/make_frames.nb.,title="fig : " ] started from the clones positions at the times indicated by the green and blue arrows in figure [ fig : clonet015 ] .
the gray - dotted regions correspond to the equilibrium md simulations of the three basins icom , icoam and fcc .
* left * : md simulations started from the stationary structures found by the clones in the ico funnel fall either back into the ico basin or in a metastable basin around @xmath148 that corresponds to an amorphous structure at the entrance of the ico funnel .
* center * : the clones starting from the stable structure found in the fcc funnel fall either back in the fcc basin , or in a faulty fcc metastable state ( blue rectangle ) or in the ico funnel , which .
both structures thus correspond to reaction paths . *
right * : position of the surface atoms of a clone that has fallen in the faulty fcc configuration after the short md .
this figure was made using a mathematica spreadsheet that can be downloaded at http://www-wales.ch.cam.ac.uk/~wales/make_frames.nb.,title="fig : " ] the clones have thus found reaction paths pointing out of their starting funnels .
the clones starting from the fcc basins find a reaction path that leads into the icosahedral funnel while the one started from the icom basin still remain in the icosahedral funnel .
this could be explained by the fact that at this temperature icom is the _
stable _ state while fcc is only metastable so that the barrier ico@xmath149fcc has to be harder to access than the one from the fcc side .
running short md starting from the clones positions reveal intermediate metastable basins , either a faulty fcc or amorphous structures .
this is the coexistence temperature between the icom and fcc minima . at such a low temperature , more and more secondary barriers
play a role so that the transition between ico and fcc becomes more and more complex . from the point of view of the time evolution of the transition
current , this means that there are more and more metastable states for the clone dynamics .
starting from the icom basin with 600 clones and @xmath136 , we once again locate the barrier between icom and liquid - like phase icoam .
at this temperature this barrier is long - lived and we do not locate the one previously found at @xmath134 that points toward the entrance of the icosahedral funnel .
starting from the fcc basin with @xmath151 , the @xmath145 clones find several barriers that constitute a multi - step reaction path toward the icosahedral funnel .
once again , the larger the friction the longer the clones spend on intermediate barriers.@xcite and @xmath142 plans at @xmath150 . the clones starting from ico
find the barrier between icom and icoam ( black symbols , @xmath152 ) .
starting from fcc , the clones find a succession of barriers that leads toward the icosahedral funnel ( green symbols at @xmath153 , blue symbols at @xmath154 and cyan symbols at @xmath155 ) . *
right : * we plot @xmath119 as a function of time for the clones starting from the icom basins ( red symbols ) and the fcc basin ( magenta symbols ) . ]
the position of the clones corresponding to the successive metastable barriers are shown in figure [ fig : clonest012 ] .
note that the typical times needed for locating the barriers are of the order of @xmath156 , that is _ seven orders of magnitude smaller _ than the reaction times between icom and fcc , which is of the order of @xmath157 .
@xcite started on the green ( left ) , blue ( center ) and black ( right ) arrows in figure [ fig : clonest012 ] .
* left : * starting from the first stationary structure found in the fcc funnel , the clones relaxes mostly in the fcc basin and in the faulty fcc configuration shown in [ fig : t015histoclones ] .
* center * the second barrier is close to the commitor between the ico and the fcc funnel : the clone population relaxes almost equally in both funnel ( @xmath158 of the clones fall back in the icosahedral funnel while @xmath159 enter the fcc funnel ) . *
right * clones started from the barrier between icom and icoam populate both basins .
note that the relaxation is much slower than for the other barrier because of the entropic nature of this barrier . ]
running short md simulations starting from the clone positions on the barriers and constructing the corresponding histograms reveals various intermediate metastable basins in the ico and fcc funnels ( see figure [ fig : t012mdafterclones ] ) .
the fact that @xmath118 and @xmath119 are not good reaction coordinates is confirmed in this figure : the first plateau ( green points on figure [ fig : clonest012 ] ) seems to be _
after _ the faulty configuration when going from the fcc funnel to the icosahedral one but the md starting from this barrier falls into the faulty configuration and the fcc basin , which seems to indicate that this barrier is a reaction state between the fcc and the faulty configuration .
there is then a second barrier between the faulty fcc and the ico funnel ( blue dots in figure [ fig : clonest012 ] ) .
a last barrier leads to the amorphous region that separates the liquid - like phase and icom minimum .
note that in these regions the md simulations are not very helpful because the transition between icom and icoam has an entropic nature so that it is difficult to relax into the basins .
the clones , however , successfully identify the barrier between these two states .
this temperature is very close to the transition between ico and liquid - like phase .
as shown by free - energy studies , the barrier between the fcc and the ico funnels is very low and the fcc basin is rather unstable .
@xcite clones starting from the fcc basin do not stabilize on any structure because there is no proper ` rare barrier ' and md simulations starting from fcc immediately falls into the icosahedral funnel .
@xcite starting @xmath161 clones from the ico basin at @xmath162 , they rapidly find a barrier connecting to the liquid - like phase .
later on , activated events lead the clones to locate a reaction path leading towards the fcc funnel . starting md from this barrier
show that the clones relax into the fcc and ico funnel .
clones are started at @xmath160 in the icom basin where they spend some time ( first first time steps after @xmath163 , green dots ) before locating the barrier toward the fcc funnel ( first five time steps after @xmath164 , blue crosses ) .
* center : * when the clones have found the barrier ( @xmath165 ) a standard md starts and relaxes as it should into the two funnels ( black dots , five first snapshots after @xmath166 ) . * right : * evolution of @xmath130 .
the cloning is stopped at @xmath165 and a normal md follows . ] as mentioned above , it is hard for the clones to stabilize because the fcc funnel is barely metastable and the barrier crossed while going from fcc to ico is rather flat at this temperature .
it is thus quite surprising that they nevertheless manage to do so while starting from the ico basin .
if one starts from the fcc funnel , the clones almost immediately fall into the ico funnel and then from there can locate the barrier again , but we were not able to stabilize the barrier when coming from the fcc basin .
this might be due to the fact that clones stabilize reactions that take place on long time - scale ( ico@xmath149fcc ) , but not short - time relaxations ( fcc@xmath149ico ) . *
right : * starting from the end point of the simulations at t=0.12 , we run a standard cloning simulation at @xmath167 . after a time @xmath168
the 600 clones are still on the structure that had localized at @xmath150 , which is thus very stable . ]
if one starts at such a low temperature from one of the various metastable basins , the clones remain trapped for a time longer than the simulation time .
one can however use a temperature annealing to locate the barriers .
if one starts the cloning simulation at @xmath150 or higher , it is quite easy , as we saw above , to localize the barriers .
the temperature can then be decreased to @xmath167 and the clones remains on the structure that were localized at a higher temperature ( see figure [ fig : annealing ] ) .
the algorithm we have discussed in this paper may be characterized as one that simulates the evolution in time of the current distribution , rather than that of the configuration . because the time for the escape current to be established is often much smaller than the passage time itself , the method is able to find the transition paths very efficiently .
the method has several attractive features : _ i _ ) it does not require any previous knowledge of the relevant reaction coordinates . on the other hand , if an approximation of the reaction path is known _ a priori _ , one may always start the clones along this path , and they will populate the true current distribution in a shorter time .
_ ii _ ) because the target of the dynamics is the reaction path distribution itself , one may perform simulated annealing in path space : first populating the reaction path corresponding to relatively high temperature , and then refining it to the lower , target temperature .
repeated annealing can also be used to locate several competing barriers in system with multiple reaction mechanisms .
_ iii _ ) the reaction current vanishes between mutually thermalized regions .
@xcite this is why at longer times , the system converges to the barriers that take longer to cross , _ irrespective of whether they are of entropic or energetic nature_. this may be an advantage in cases in which the energy landscape is not in itself dominant , but rather the multiplicity of paths dominates . _ iv
_ ) the construction of the transition current and the cloning algorithm also applies for non - equilibrium systems where the forces derive locally , but not globally from a potential , such as a system with leads at the edges having a potential difference .
reaction paths between non - equilibrium metastable states , which can not be described in term of a free energy , may be studied in the same way .
the only difference is that the average does not vanish in the long time limit and converges instead to the steady - state transition current .
note that since the method does not require the knowledge of the reaction coordinate , it could be used efficiently in systems with competing reactions where one does not know _ a priori _ the end points of the reaction paths .
this would for instance be particularly interesting when studying the crystallization of suspensions of oppositely charged colloids .
@xcite in principle , the reaction time may be expressed directly in terms of the ( unnormalized ) reaction current .
it may also be recovered from the weights carried by the clones , which may possibly be achieved from importance sampling in a lyapunov - weighted ensemble of trajectories .
@xcite however , the method , as it stands does not allow one to calculate the reaction _ time _ with great precision , due to the exponential nature of the timescale . further work is required in this direction .
acknowledgement : we thank the esf programme `` simbioma - molecular simulations in biosystems and material science''(m.p . ) and epsrc ep / h027254 ( j.t . ) for funding , d. wales and c. valeriani for repeated fruitful discussions and f. calvo and d. wales for giving us several numerical routines .
we report here the derivation of the time evolution equation for the transition current in the underdamped ( i.e. , kramers ) case of section [ sec : kram ] .
the classical kramers probability current @xmath169 , presented in equations , can indeed be written , as in the fokker - planck case ( eq . ) , in the operatorial form @xmath170 , where the components of the current kramers operator are @xmath171 in equations of section [ sec : kram ] , a _ transition current _ @xmath172 has been introduced , which can be expressed in turn with operators as @xmath173 where @xmath174 is the kramers current operator reported above , giving the usual _ phase - space _
current @xmath169 , and the ` transition ' operator @xmath175 @xmath176 is a divergenceless term . as already remarked in section [ sec : kram ] , the transition current still satisfies the continuity equation @xmath177 thanks to the divergenceless of @xmath176 .
we have introduced here the phase - space divergence @xmath178 as in section [ sec : fp ] , we proceed now in deriving explicitly the time evolution equation of @xmath172 .
multiplying both sides of the continuity equation above ( indeed identical to the kramers equation ) by the transition current operator leads to @xmath179 on the l.h.s .
the transition current operator can be commuted with the time derivative . the r.h.s .
of can be rewritten with commutators as @xmath180+\left[\hat{\mathbf{t}},\nabla_{\mathbf{r},\mathbf{v}}\cdot\hat{\mathbf{j}}\right]\ ] ] using and the zero divergence property of @xmath181 .
resorting to definitions of the current operator @xmath182 and the transition operator @xmath181 given in , and , explicit expressions for commutators in can be recovered with straightforward algebra : the term @xmath183 $ ] gives [ a5 ] @xmath184p & = -\sum_{i}\delta_{ia}\left(j_{v_{i}}-\frac{\gamma}{\beta m_i}\partial_{v_{i}}\right)p\\ \left[\hat{j}_{v_{a}},\nabla_{\mathbf{r},\mathbf{v}}\cdot\hat{\mathbf{j}}\right]p & = \sum_{i}\delta_{ia}\left(\gamma j_{v_{i}}-\frac{\gamma}{\beta m_i}\partial_{r_{i}}\right)p-\sum_{i}\frac{1}{m_i}\frac{\partial^{2}v}{\partial r_{i}\partial r_{a}}v_{i}p\end{aligned}\ ] ] while @xmath185 $ ] can be expressed as [ a7 ] @xmath186p&=\sum_{i}(\beta m_i)^{-1}\delta_{ia}\left(\partial_{r_{i}}+\gamma\partial_{v_{i}}\right)p\\ \left[\hat{t}_{v_{a}},\nabla_{\mathbf{r},\mathbf{v}}\cdot\hat{\mathbf{j}}\right]p&=\sum_{i}(\beta { m_i}^2)^{-1}\frac{\partial^{2}v}{\partial r_{i}\partial r_{a}}\partial_{v_{i}}p\end{aligned}\ ] ] inserting now and in the r.h.s .
of eq . yields [ a8 ] @xmath187 that can be recasted as @xmath188 leading to equation .
in our study of the lj38 system , we carried out simulations of the stochastic dynamics of the clones but did not explicitly make the averages that would yield the transition current .
we argued in section [ sec : evolcuralgo ] that if one only looks for the reaction paths , then these averages are not necessary . in this appendix
, we illustrate this claim on a simple 1d example that also allows us to discuss several aspects of the clone dynamics , namely the metastability , the finiteness of the clone population and the role of the initial condition .
we thus consider a system undergoing an underdamped langevin dynamics @xmath189 where @xmath190 is a potential with two barriers , plotted in figure [ fig : potentiel ] .
we ran the clone dynamics for 2000 clones starting in the left well and carried out the averages . .
* right * : the green crosses are the position of the 2000 clones after a time @xmath191 .
the black arrows correspond to the averages and indeed point tangentially to the reaction path .
the color code and contour lines corresponds to the value of the hamiltonian @xmath192 to do so , we constructed an approximate density from the positions and vectors @xmath193 of each of the @xmath194 clones : @xmath195 in principle , the @xmath196 should be dirac functions but for practical purposes we replaced the one acting on the phase space coordinate by the bell - shaped function @xmath197 where @xmath198 and @xmath199 is a normalization constant . finally , using and in ,
we construct the transition current from the numerical data by computing @xmath200 on a grid every @xmath201 and plot the resulting vector if its norm is larger than @xmath202 . for visualization purposes ,
we plot in the figures the vectors @xmath203 times longer than they really are .
we started a simulation with @xmath204 clones in the left well , around @xmath205 , with unitary vectors @xmath206 pointing at random .
the temperature is set to @xmath207 and the friction to @xmath208 so that the mean first passage time across the barrier is @xcite @xmath209 where @xmath210 is the first maximum of the potential @xmath211 .
the results of the simulation after a time @xmath191 are plotted in figure [ fig : potentiel ] .
the clones have already populated the barrier . @xcite
as can be seen , the averages along the reaction path are non - zero and result in vectors tangent to the reaction path , pointing toward the left well . at later times ,
two processes take place , roughly on the same time scale .
firstly , more and more clones come back from the central well to the left one .
their vectors @xmath59 are always tangent to their trajectories , but can be pointing toward the left or the central well with equal probability . indeed ,
if @xmath212 is a possible trajectory of the system , then so is @xmath213 . as a result
, the averages may cancel out at large times , when the subpopulations of clones whose vectors @xmath214 point toward the central and the left wells balance .
this is how numerically the transition current is supposed to vanish at large time ( another possibility being that all the clones leave a region of phase - space , because of finite population - size effects ) .
secondly , some clones reach the barrier leading to the right well and duplicate , which results in populating the second reaction path . since the clones did not have time to fall in the right well and cross back the barrier towards the central well with vectors @xmath59 pointing in the opposite direction , the average does not cancel along this reaction path .
both effects can be seen in figure [ fig : clones1dlater ] : the clones populate both barriers ; the average cancels out over the first barrier but not yet over the second one .
this shows that the clone dynamics do locate the barriers and remain on the reaction paths even though the transition current may average out when the two wells separating a barrier equilibrate .
this thus validates our approach to locate the reaction paths in the lj38 system .
note that if one wishes to study quantitatively the transition current , two modifications would need to be done to our algorithm .
first , the initial condition should not be taken at random but constructed as proposed in section [ sec : fp ] .
second , rather than simulating all the clones in parallel while maintaining their population constant , it may be advantageous to run them sequentially , starting one run for every offspring of every clones , as is done for instance with the ` go with the winner ' methods .
@xcite the constrain on the total population being fixed indeed affects the metastability of the clone dynamics and increases finite size effects . for instance , if there are @xmath215 and @xmath216 clones on the same reaction path , with vectors pointing in opposite directions , both populations grow exponentially with the same rate .
now , if the total population is kept constant , then the smallest sub - population disappears on average exponentially fast . using a
` go with the winner ' method would palliate this drawback but would result in additional computational costs difficult to estimate beforehand .
note however that the width of the clones distribution in the configuration space is of order @xmath218 @xcite so that for very large friction , these clouds start to cover several barriers at the same time and their dynamics can then change . | reaction paths and probabilities are inferred , in a usual monte carlo or molecular dynamic simulation , directly from the evolution of the positions of the particles .
the process becomes time - consuming in many interesting cases in which the transition probabilities are small .
a radically different approach consists of setting up a computation scheme where the object whose time evolution is simulated is the transition current itself .
the relevant timescale for such a computation is the one needed for the transition probability _ rate _ to reach a stationary level , and this is usually substantially shorter than the passage time of an individual system . as an example , we show , in the context of the ` benchmark ' case of 38 particles interacting via the lennard - jones potential ( ` lj@xmath0 ' cluster ) , how this method may be used to explore the reactions that take place between different phases , recovering efficiently known results and uncovering new ones with small computational effort . |
long - range correlations in a fluid far from equilibrium is one of the most salient features of nonequilibrium fluctuations @xcite . since the seminal discovery of the @xmath8 enhancement of the rayleigh component of the structure factor in a fluid with an uniform temperature gradient by kirkpatrick , cohen and dorfman @xcite , where @xmath1 is
the wave number , algebraic decays of correlation functions have been found theoretically in a wide class of nonequilibrium systems @xcite .
the existence of such generic long - range correlations also have been confirmed in a series of the detailed light scattering experiments in pure liquids @xcite and in binary mixtures @xcite .
although the agreement between the theoretical studies and the experimental ones are sufficiently quantitative , it have been limited to relatively high wave number regions . on the other hand , large - scale , long - time behavior of such nonequilibrium fluctuations is attracting much more attentions in recent years @xcite .
ultra - low - angle light scattering experiment performed by vailati and giglio @xcite have revealed an impressive gravity - induced quench of the @xmath8 divergence of the rayleigh line intensity at very small @xmath1 range @xcite in a binary mixture subjected to an uniform concentration gradient driven by the large soret effect .
so far as we know , this is the first experimental demonstration that true asymptotic behavior of a long - range correlation at large distances is qualitatively different from that predicted by the linear response theory at short distances .
however , from the theoretical point of view , the gravity - induced effect seems to be somewhat exceptional because it can be understood within the simple linear response theory . generally speaking ,
a full nonlinear analysis of hydrodynamic equations is required for studying large - scale transport and fluctuation properties of a fluid @xcite . therefore , mainly because of its technical difficulty
, studies beyond linear responses are rare in systems far from equilibrium so far .
only one exception is a system undergoing uniform shear flow where effects of the shear advection can be treated in an analytic way @xcite .
it has been predicted theoretically that the density autocorrelation function and the density - momentum correlation function in a sheared compressible fluid exhibit much weaker divergent behavior in small wave number limits than the @xmath8 divergence found in large wave numbers @xcite .
a similar result has also been predicted for an incompressible fluid @xcite .
nevertheless , it might be difficult to verify these predictions directly by using available experimental techniques , because the crossover length scales of these systems are so macroscopic .
the purpose of this paper is to emphasize effects of the shear upon long - range correlations which can allow an experimental confirmation . to this end
, we consider the particularly simple system , a ( dilute ) low - molecular weight binary mixture with imposed concentration and uniform shear gradients @xcite . on the basis of fluctuating hydrodynamics approach @xcite
, we calculate an static concentration autocorrelation function which contributes to the total intensity of the rayleigh peak in a light scattering experiment . in sec .
[ sec : model ] , hydrodynamic equations for the concentration and the velocity field are introduced with suitable random forces .
we obtain general forms of space - time correlation functions of the hydrodynamic variables by solving the linearized equations around the steady state under imposed boundary conditions . in sec .
[ sec : long ] , the static correlation function of concentration fluctuations is expressed in terms of a universal function which describes a shear - induced frustration of the long - range correlation . this function
is numerically evaluated and is compared with asymptotic functions that are analytically derived . in sec .
[ sec : discussion ] , we provide a simple physical interpretation of our results .
we also present rough estimations of the crossover length scales for a realistic fluid to suggest possible experimental verifications .
our conclusion is given in the last section .
the mass diffusion flux @xmath9 generally depends not only on the concentration gradient @xmath10 but also on the temperature gradient @xmath11 and on the pressure gradient @xmath12 @xcite .
however , when the mass concentration @xmath4 is small , the thermal diffusion ratio is expected to be rather small because it must vanish in a pure liquid @xcite .
we thus can safely neglect the term proportional to @xmath11 , as well as the term proportional to @xmath12 in the expression of @xmath9 because we do not consider any gravity - induced effect . to make the analysis below as simple as possible , we further assume that the mixture is incompressible . consequently , the only relevant hydrodynamic variables are the concentration @xmath4 and the momentum @xmath13 of the mixture .
the equations describing the time evolution of these variables supplemented by the suitable random forces are of the form @xcite @xmath14 where the diffusion flux is given as @xmath15 , and the hydrostatic pressure @xmath16 is determined from the incompressibility condition @xmath17 . here @xmath6 is the mass diffusive coefficient , @xmath18 is the kinematic viscosity of the mixture ( @xmath19 the zero shear viscosity ) . the random forces @xmath20 and @xmath21 are the random concentration flux and the random stress tensor , respectively .
the correlations of these random forces retain their local equilibrium values given by @xcite @xmath22 \delta^3({\bf r}-{\bf r } ' ) \delta(t - t ' ) , \label{fdt - v01}\end{aligned}\ ] ] and @xmath20 is uncorrelated with @xmath21 . in the following , we assume that the osmotic compressibility @xmath23 and the mass diffusivity @xmath6 are both independent of the concentration @xmath4 , where @xmath24 is the chemical potential of the mixture @xcite . this assumption may be valid when the gradient @xmath25 is not sufficiently large .
we note that the local equilibrium assumption in eqs .
( [ fdt - c01 ] ) and ( [ fdt - v01 ] ) is justified from the fact that the random forces represent the fast , and localized molecular process which can not be affected by the macroscopic gradients of the hydrodynamic quantities .
although we are primarily interested in the bulk fluctuation properties of the mixture , appropriate boundary conditions are required to specify a macroscopic steady state of the averaged thermodynamic quantities @xcite .
we adopt here the standard lees - edwards type boundary conditions @xcite , which can be written explicitly as @xmath26 and @xmath27 where @xmath28 is the linear system size of @xmath29 direction that is parallel to both the concentration and shear gradient directions .
it is also assumed that the system is infinitely large in all other directions .
the statistically averaged , time - independent equations ( [ eq - c01 ] ) and ( [ eq - v01 ] ) under the imposed boundary conditions ( [ bc - v ] ) and ( [ bc - c ] ) have solutions @xmath30 where @xmath31 stands for the concentration in the @xmath32 plane . we here chosen our coordinate frame
whose origin is at rest by making use of the galilean invariance .
it is also easy to show that , within the present model , the average concentration profile ( [ avsol - c ] ) is linearly stable for a perturbation with any wave number under the imposed velocity profile ( [ avsol - v ] ) @xcite .
now let @xmath33 and @xmath34 in eqs .
( [ eq - c01 ] ) and ( [ eq - v01 ] ) .
we then obtain the linearized equations for fluctuations @xmath35 and @xmath36 as @xmath37 where an equation for the pressure fluctuation follows from the divergence of the equation for @xmath38 : @xmath39 here we introduced the random variables @xmath40 and @xmath41 .
equation ( [ eq - c02 ] ) implies that the only @xmath29 component of the velocity fluctuation @xmath42 can affect the dynamics of @xmath35 . eliminating @xmath43 from eq .
( [ eq - v02 ] ) by using eq .
( [ press ] ) , we obtain the linearized hydrodynamic equations relevant to the present purpose of the form @xmath44 where the fourier transform of an arbitrary function @xmath45 is defined as @xmath46 note that the random variable in eq .
( [ eq - v03 ] ) is re - defined by @xmath47 .
this variable and @xmath48 satisfy the correlation properties given from eq .
( [ fdt - c01 ] ) and eq .
( [ fdt - v01 ] ) as @xmath49 and @xmath50 , where @xmath51 .
the set of equations ( [ eq - c03 ] ) and ( [ eq - v03 ] ) are most easily solved by making a transformation to the local lagrangian coordinate given by @xmath52 @xcite . applying the standard manipulation described elsewhere @xcite
, we readily find the solutions of the form @xmath53 where @xmath54 and @xmath55 here @xmath56 is the time - dependent wave vector defined as @xmath57 and the green s functions are given by @xmath58 where @xmath59 and @xmath60 .
the space - time correlation function of the hydrodynamic fluctuations under the uniform shear is defined by @xcite @xmath61 for @xmath62 , where the angle bracket represents the statistical average with respect to the random variables @xmath63 and @xmath64 .
the function @xmath65 represents either @xmath35 or @xmath42 . under the uniform shear condition , the system is invariant with respect to the temporal transformation , whereas it is not invariant with respect to the spatial transformations @xcite .
the time invariance property imposes the symmetry @xmath66 this relation implies that it is sufficient to consider the function @xmath67 for @xmath68 in the following calculations , without loss of generality @xcite . substituting eqs .
( [ sol - v01 ] ) and ( [ sol - c01 ] ) into eqs .
( [ cf01 ] ) and using the properties of the random variables ( [ fdt - c02 ] ) and ( [ fdt - v02 ] ) , we can derive the time correlation functions of the hydrodynamic modes .
the calculations are somewhat lengthy , so we here give only the results : @xmath69 where @xmath70 , \label{cf - v02}\ ] ] @xmath71 , \label{cf - c02}\ ] ] and the cross - correlation function given by @xmath72 here the green s functions are written explicity as @xmath73 with @xmath74 where @xmath75 . in the derivations of eqs .
( [ cf - v02])-([cf - cv02 ] ) , we have made use of the identity @xmath76 .
although eqs .
( [ cf - v02])-([cf - cv02 ] ) are rather complicated , the equal - time correlation functions have somewhat simpler forms because of the recovery of the translational invariance with respect to the space . setting @xmath77 in eq .
( [ cf - v02])-([cf - cv02 ] ) , we obtain @xmath78 , \label{cf - v03}\ ] ] @xmath79 , \label{cf - c03}\ ] ] and @xmath80 equations ( [ cf - v03])-([cf - cv03 ] ) together with eqs .
( [ t ] ) and ( [ tau ] ) are the main results of this section . the same result as eq .
( [ cf - v03 ] ) has also been derived for the simple fluid in ref .
thus the correlation function between the momentum fluctuations is unaffected by the steady concentration gradient within the present analysis .
we shall first examine the two limiting cases studied previously in eqs .
( [ cf - v03])-([cf - cv03 ] ) . at zero shear rate ,
( [ cf - v03])-([cf - cv03 ] ) are reduced to the form @xmath81\nonumber\\ & = & \frac{k_bt}{\rho}\chi_c\left(1+\frac{|\nabla c|^2 } { \chi_cd(\nu+d)}\frac{|\hat{{\bf k}}_{\perp}|^2}{k^4}\right ) , \label{cf - c04}\\ c_{cv}({\bf k } ) & \rightarrow & -2\frac{k_bt}{\rho } |\nabla c|(\nu k_{\perp}^2)\int_0^{\infty}dt e^{(\nu - d)k^2t}\int_t^{\infty}dt ' e^{-2\nu k^2 t ' } = -\frac{k_bt}{\rho}\frac{|\nabla c|}{(\nu+d ) } \frac{|\hat{{\bf k}}_{\perp}|^2}{k^2}. \label{cf - cv04}\end{aligned}\ ] ] equation ( [ cf - v04 ] ) is the standard result in thermal equilibrium @xcite .
in contrast to this , the second term in eq .
( [ cf - c04 ] ) and eq .
( [ cf - cv04 ] ) show the anomalous enhancements of the hydrodynamic fluctuations in the presence of the concentration gradient .
these results are first predicted by kirkpatrick _
et al_. using mode coupling and kinetic theory @xcite and subsequently confirmed by ronis and procaccia using fluctuating hydrodynamics @xcite . on the other hand , in the limit of a small concentration gradient , eqs .
( [ cf - c04 ] ) and ( [ cf - cv04 ] ) converge to the equilibrium forms given by @xmath82 and @xmath83 regardless of the presence of the shear flow .
however , the momentum - momentum correlation function becomes long - ranged in this case because of the influence of the shear @xcite .
two asymptotic forms in the selected direction of the wave vector @xmath84 are given by @xcite @xmath85 for @xmath86 , and @xmath87 for @xmath88 , where @xmath89 is the gamma function .
the direct numerical evaluation of eq .
( [ cf - v03 ] ) is also shown in fig .
[ fig : v - v corr ] .
the length scale which characterize this crossover is introduced as @xmath90 plots of numerically evaluated scaling function @xmath91 , where the momentum autocorrelation function is written in the form @xmath92 $ ] .
the dashed and dotted lines represent the aymptotic functions for large @xmath93 ( eq .
( [ cf - v05a ] ) ) and for small @xmath94 ( eq .
( [ cf - v05b ] ) ) , respectively.,width=302 ] in the following , we shall focus our attention to the behavior of eq .
( [ cf - c03 ] ) for arbitrary magnitudes of the shear rate @xmath5 and the concentration gradient @xmath25 .
it is easy to show that eq .
( [ cf - c03 ] ) can be written in the form @xmath95 we here consider the case where the scattering wave vector @xmath96 is perpendicular to the concentration gradient @xmath10 , i.e. , @xmath97 .
this scattering geometry still holds all of the essential features of the nonequilibrium effects found in this system , and is actually the configuration adopted in most of the previous light scattering experiments @xcite .
the scaling function @xmath98 in eq .
( [ cf - c05 ] ) in this geometry is given by @xmath99 where @xmath100 and @xmath101 is the ratio of the mass diffusivity to the viscous diffusivity , which is usually much smaller than the unity .
as can be understood from eq .
( [ f01 ] ) , the effect of the shear is most likely to become evident in the particular direction of the wave vector given by @xmath84 . to simplify the analysis , we restrict our interest to the fluctuations in this direction . because of this simplification ,
the asymptotic behavior for small @xmath102 limit is easily found to be @xmath103 equation ( [ f - as01 ] ) suggests that the enhancement of the nonequilibrium fluctuations is severely restricted in the long wavelength limit . on the other hand , the opposite asymptotic region , i.e. , the short wavelength limit ,
is represented by @xmath104 , where the second characteristic length scale @xmath105 is given by @xmath106 it turns out that the asymptotic behavior of @xmath107 for @xmath104 has the same wave number dependence given by the second term in eq .
( [ cf - c04 ] ) .
this implies that thermal fluctuations of the hydrodynamic variables for @xmath104 are little affected by the shear and can dissipate purely thermally .
consequently , the static concentration autocorrelation function in the two limiting cases of short and long wave numbers are written as @xmath108 for @xmath86 , and @xmath109 for @xmath110 , where the numerical constant @xmath111 is given by @xmath112 .
the scaling function @xmath107 as a function of @xmath102 obtained by the numerical integration of eq .
( [ f01 ] ) for @xmath113 and @xmath114 .
the lines are the asymptotic results derived analytically .
note that the crossover wave numbers given by @xmath115 and @xmath116 correspond to @xmath117 and @xmath118 , respectively.,width=321 ] in fig .
[ fig : c - c corr ] , a numerically calculated scaling function @xmath119 is plotted with the asymptotic functions for large and small @xmath120 deduced from eqs .
( [ cf - c06a ] ) and ( [ cf - c06b ] ) , respectively .
although the intermediate region given by @xmath121 seems to have @xmath122 dependence , whether it is a true scaling regime or just a crossover is still unclear at present .
note , however , that this region is practically extended in a wide range of the wave numbers , which is typically given as @xmath123 due to the large asymmetry between the magnitude of the viscous and mass diffusivity .
therefore , we expect that this unclearness will be clarified by evaluating @xmath119 for a much smaller @xmath124 than that considered here ( @xmath125 ) , though it can not be done in the present study primarily because of the limitation of the numerical accuracy .
as pointed out in a number of literature , the long - range correlation is originated from the coupling of momentum fluctuations with concentration fluctuations through a macroscopic concentration gradient @xcite .
this is sometimes called a linear mode - coupling effect @xcite . following the discussions given in ref .
@xcite , let us suppose a small fluid element of linear size @xmath126 in the mixture in the absence of the shear .
this fluid portion is carried along the direction of the gradient by a spontaneously generated momentum fluctuation of its lifetime given by @xmath127 . because the surrounding fluid has a different average concentration , the large concentration difference persists for a time @xmath128 as it relaxes purely diffusively . considering that @xmath129
, one can notice that a short - living , spontaneous momentum fluctuation can creates a long - lasting , large - amplitude concentration fluctuation , which results in a long - range spatial correlation between concentration fluctuations in the steady state .
however , shear flow limits the size of such fluctuations significantly , because a sufficiently large fluctuation is drawn out and is even broken up by the shear before it disappears by the diffusion @xcite . in our system ,
@xmath105 and @xmath130 correspond to the crossover length scale from the diffusion - dominated decay to the shear - dominated decay of concentration and momentum fluctuations , respectively . because @xmath105 is much smaller than @xmath130 in practice ,
the overall behavior of @xmath131 is expected to have three distinctive wave number domains .
they can be sketched as follows .
\(i ) when a small portion of mixture of linear size larger than @xmath130 is carried along the direction of the gradient , it undergoes a significant shear deformation and becomes highly anisotropic . because the lifetime of the concentration fluctuation is controlled by the shortest length scale of its spatial extent , this fluctuation dissipates thermally much faster than that in a quiescent fluid .
thus the long - range correlation between concentration fluctuations is severely suppressed by the shear flow in this wave number region .
\(ii ) when @xmath132 , a momentum fluctuation can displace a small parcel of fluid of size @xmath126 without a notable shear deformation .
however , if @xmath133 , the shear flow strongly affects the decay of this large - amplitude concentration fluctuation before it dissipates thermally .
then the decay of the fluctuation is still faster than that in a quiescent fluid in this region , and the resulting correlation exhibits a weaker divergent dependence on the wave numbers @xmath1 than @xmath8 .
\(iii ) when a size of a fluctuation is much smaller than @xmath105 , the effect of the shear becomes rather weak because a thermal diffusive decay is faster than the shear deformation time scale ; @xmath134 .
therefore the usual linear response argument for a quiescent fluid presented in the beginning of this section is fit for this regime .
figure [ fig : c - c corr ] shows that the deviation from the @xmath8 divergence becomes pronounced at the wave number @xmath135 .
it should be emphasized that the presence of @xmath136 is the clear evidence of the shear - induced suppression of the long - range correlation .
although the shear - controlled asymptote is confirmed only in the wave numbers smaller than @xmath137 , it may be impossible to verify this scaling experimentally because the length scale @xmath130 is so macroscopic for relevant fluid parameters and experimentally feasible shear rates @xcite .
( for example , we can obtain @xmath138 mm@xmath139 at most , for a typical value of the viscosity @xmath140 @xmath141s@xmath139 and for a very high rate of the shear that is of the order of @xmath142 s@xmath139 . )
contrary to this situation , a typical value of the diffusion coefficient @xmath143 @xmath141s@xmath139 gives @xmath144 cm@xmath139 even for a very small rate of the shear as @xmath146 s@xmath139 .
therefore , the wave number @xmath136 , which characterizes the onset of the shear - induced frustration of the long - range correlation , is well covered by conventional low - angle light scattering methods , as well as the recently reported new optical technique @xcite . there have been developed some experimental techniques to produce a macroscopic concentration gradient in a fluid , such as utilizing a diffusive remixing process of a mixture @xcite or making use of a large soret effect driven by a steady temperature gradient @xcite .
although it is far beyond our ability to guess about , the latter method might be more favorable with respect to the compatibility with the shear flow , as well as the spatial uniformity and the temporal stationarity of the concentration gradient .
when the large soret effect is expected , however , we have to take a temperature fluctuation into account as well , because the cross - correlation between concentration and temperature fluctuation makes an important contribution to the rayleigh scattering @xcite . in that case
, the third length scale @xmath147 may enter into the theory , where @xmath148 is the thermal diffusivity of the mixture .
nevertheless , since @xmath149 for most of relevant mixtures , @xmath105 is likely to be well separated from the other two length scales .
then , so far as we are concerned with the shear - induced effect , especially with the onset of the crossover , we can expect that the present analysis is applicable to such cases .
in this paper , using fluctuating hydrodynamics , we have studied the fluctuation properties in a ( dilute ) binary mixture with imposed uniform concentration and shear gradients .
it is demonstrated that the static concentration autocorrelation function has the three distinctive wave number regions .
the length scales which characterize the crossovers between different regimes are given as @xmath150 and @xmath151 , respectively .
in particular , shear strongly suppresses the @xmath8 divergence for @xmath152 , while the asymptotic shear - controlled behavior is found analytically in the smallest wave numbers satisfying @xmath88 .
it is worth noting that the onset of the shear - induced quench of the long - range correlation may be experimentally observable .
such experimental verification will also become an important test for the applicability of fluctuating hydrodynamics to a fluid in a coupled - nonequilibrium steady state .
the author is grateful to s - i .
sasa for guiding him in the problem of nonequilibrium fluctuations , and a number of valuable discussions .
the author also acknowledges r. cerbino and a. vailati for introducing their original works and illuminating discussions which gave a motive for the present study at the visit in university of milan .
thanks are due to t. ohta and t. shibata for their hospitality in hiroshima university , where much of this work has been carried out .
lastly , the author would like to thank m. sano for his encouragements .
this work was supported by grant - in - aid for jsps fellows for young scientists , from ministry of education , culture , sports , science , and technology , japan .
99 r. schmitz , phys .
* 171*,1 ( 1988 )
. j. r. dorfman , t. kirkpatric , and j. v. sengers , annu .
phys . chem . * 45 * , 213 ( 1994 ) .
d. ronis , i. procaccia and i. oppenheim , phys .
rev . a * 19 * 1324 ( 1979 ) .
d. ronis , i. procaccia and j. machta , phys . rev . a * 22 * 714 ( 1980 ) .
j. machta , i oppenheim and i. procaccia , phys .
a * 22 * , 2809 ( 1980 ) .
s. tremblay , m. arai , and e. d. siggia , phys .
a * 23 * , 1451 ( 1981 ) .
t. r. kirkpatrick , e. g. d. cohen , and j. r. dorfman , phys . rev . a * 26 * , 950 ( 1982 ) ;
* 26 * , 972 ( 1982);*26 * , 995 ( 1982 ) .
d. ronis and i. procaccia , phys .
a * 26 * , 1812 ( 1982 ) .
b. m. law and j. c. nieuwoudt , phys .
a * 40 * , 3880 ( 1989 ) .
j. lutsko and j. w. dufty , phys .
a * 32 * , 3040 ( 1985 ) .
a. l. garcia , m. m. mansour , g. c. lie , m mareschal , and e clementi , phys .
a * 36 * , 4348 ( 1987 ) .
a. vailati and m. giglio , phys .
e * 58 * , 4361 ( 1998 ) .
p. l. garrido , j. l. lebowitz , c. maes , and h. spohn , phys .
a * 42 * , 1954 ( 1990 ) .
m. yoshimura and t. r. kirkpatrick , phys .
b * 54 * , 7109 ( 1996 ) . b. m. law , r. w. gammon , and j. v. sengers , phys
lett . * 60 * , 1554 ( 1988 ) .
p. n. segre , r. w. gammon , j. v. sengers , and b. m. law , phys .
a * 45 * , 714 ( 1992 ) .
p. n. segre , r. w. gammon , and j. v. sengers , phys .
e * 47 * , 1026 ( 1993 ) .
it is worth noting that such a motivation comes in part from the insight on the possible relationship between a non - analytic dependence of a thermodynamic quantity on a nonequilibrium control parameter and a long - range correlation of its fluctuation .
see also ref .
a. vailati and m. giglio , phys .
lett . * 77 * , 1484 ( 1996 ) ; nature ( london ) * 390 * , 262 ( 1997 ) .
p. n. segre , r. schmitz , and j. v. sengers , physica ( amsterdam ) * 204a * , 399 ( 1994 ) .
d. forster , d. r. nelson and m. j. stephen , phys .
rev . a * 16*,732 ( 1977 ) .
a. onuki , phys .
a * 70 * , 31 ( 1979 ) .
a. onuki and k. kawasaki , ann .
* 121 * , 456 ( 1979 ) .
j. w. dufty , phys .
a * 30 * , 1465 ( 1984 ) .
j. f. lutsko and j. w. dufty , phys .
e * 66 * , 041206 ( 2002 ) .
h. wada and s. sasa , phys .
e * 67 * , 065302(r ) ( 2003 ) .
to our knowledge , there has been only one theoretical study of a system similar to ours [ c. tremblay and a .- m .
s. tremblay , phys .
a * 25 * , 1692 ( 1982 ) ] and no experimental report available .
this study was concerned with the brillouin lines for a pure liquid subjected to both a temperature gradient and a shear flow .
l. d. landau and e. m. lifshitz , _ fluid mechanics _ ( pergamon , newyork , 1975 )
. s. r. de groot and p. mazur , _ non - equilibrium thermodynamics _ ( north - holland , amsterdam , 1962 ) .
d. brogioli , a. vailati , phys .
e * 63 * , 012105 ( 2000 ) .
a. lees and s. edwards , j. phys .
c * 5 * , 1921 ( 1972 ) .
note that the full nonlinear hydrodynamic equations generally have a long wavelength instability in uniform shear flow .
see , m. lee , j. w. dufty , j. m. montanero , a. santos , and j. f. lutsko , phys .
76 * , 2702 ( 1996 ) . however , the critical wavelength scale for the stability is quite large for a moderate rate of the shear considered here .
d. forster , _ hydrodynamic fluctuations , broken symmetery , and correlation functions _ , ( perseus books , 1990 ) .
d. a. weitz , nature ( london ) * 390 * , 233 ( 1997 ) .
d. brogioli , a. vailati and m. giglio , phys .
e * 61 * , r1 ( 2000 ) .
d. brogioli , a. vailati and m. giglio , europhys . lett . * 63 * , 220 ( 2003 ) .
r. cerbino , a. vailati and m. giglio , phys .
e * 66 * , 055301(r ) ( 2002 ) . | a static correlation function of concentration fluctuations in a ( dilute ) binary liquid mixture subjected to both a concentration gradient and uniform shear flow is investigated within the framework of fluctuating hydrodynamics .
it is shown that a well - known @xmath0 long - range correlation at large wave numbers @xmath1 crosses over to a weaker divergent one for wave numbers satisfying @xmath2 , while an asymptotic shear - controlled power - law dependence is confirmed at much smaller wave numbers given by @xmath3 , where @xmath4 , @xmath5 , @xmath6 and @xmath7 are the mass concentration , the rate of the shear , the mass diffusivity and the kinematic viscosity of the mixture , respectively
. the result will provide for the first time the possibility to observe the shear - induced suppression of a long - range correlation experimentally by using , for example , a low - angle light scattering technique . |
non - linear sigma models in 1 + 1 dimensions play an important role in several areas of theoretical and mathematical physics , see e.g. @xcite for a review .
they become accessible to analytical methods in particular when the target manifold is a coset manifold ( maximally symmetric ) , such as @xmath4 .
this model by construction has a manifest internal @xmath2-symmetry as well as further hidden symmetries that make it integrable .
its target space is the riemannian manifold @xmath5 , and its internal symmetry group is compact .
string theory is closely related to non - linear sigma models , where the target space comes into play as the spacetime in which the strings propagate .
often , it is assumed to be of the form @xmath6 , where @xmath7 is a suitable compact riemannian manifold representing the @xmath8 extra dimensions to have certain well - known special values , depending whether or not one includes fermions .
we will not consider such symmetries in this paper , so in this sense the connection to string theory is not a direct one . ] .
however , there is no a priori reason not to consider more general target spaces where not just the extra dimensions are curved .
for instance , it seems natural to consider target spaces such as desitter spacetime @xmath9 , which , like the sphere , is a coset manifold ( maximally symmetric space ) .
this type of non - linear sigma model would be expected to have an internal @xmath1-symmetry ( which is not compact ) together , perhaps , with further hidden symmetries by analogy with the @xmath2 non - linear sigma model . to exploit the hidden symmetries , say , in the @xmath2-model
, one may take advantage of the fact that its scattering matrix must be factorizing . in combination with the internal @xmath2 symmetry , natural assumptions about the `` particle spectrum '' ( basically the representation of @xmath2 ) , hints from perturbation theory , and the highly constraining relations imposed by crossing symmetry , yang - baxter relation , analyticity , etc .
, one can often guess the form of the 2-body scattering matrix , which then consistently determines the @xmath8-body scattering matrix @xcite . in order to derive from such a scattering matrix quantities associated with the local operators of the theory
, one can for example follow the bootstrap - form factor program @xcite .
the aim of this program is to determine the matrix elements of local operators between in- and out - states ( form factors ) , which are found using the scattering matrix and various a priori assumptions about the form factors .
the program is largely successful but runs into technical difficulties when attempting to compute higher correlation functions in terms of the form factors , or , indeed , when trying to even show that the corresponding series converge .
an alternative approach is to construct the operator algebras generated by the local quantum fields by abstract methods ( see @xcite , and @xcite for a review ) .
the input is again the scattering matrix , but the procedure is rather different .
first , one constructs certain half - local generator fields .
these `` left local fields '' @xmath10 play an auxiliary role and are constructed in such a way that with each of them , there is an associated `` right local field '' @xmath11 such that @xmath12=0 $ ] if @xmath13 and @xmath14 are space like related points in 1 + 1 dimensional minkowski space _ and _ @xmath14 is to the right of @xmath13 in a relativistic sense .
we will actually work with a similar construction for a `` chiral half '' of a massless theory on a lightray , where @xmath15 = 0 $ ] if @xmath16 with @xmath17 lightray coordinates @xcite .
both on two - dimensional minkowski space and in the chiral lightray setting , the left and right local fields generate left and right local operator algebras , and suitable intersections of these algebras then contain the truly local fields @xcite .
the latter are thereby characterized rather indirectly , and indeed , the local fields do not have a simple expression in terms of the auxiliary semi - local objects , but rather reproduce the full complexity of the form factor expansion @xcite . in this article
, we will show that such constructions also work for non - compact target spaces such as @xmath18 .
since the internal symmetry group , @xmath1 , is non - compact , its non - trivial unitary representations must necessarily be infinite - dimensional .
this is an obvious major difference , say , to the @xmath2-model , where the basic representation under which the single particle states transform is the fundamental representation , which is finite ( @xmath19- ) dimensional . despite this difference
, one may proceed and ask whether the algebraic method can be generalized to non - compact groups such as @xmath1 . for this ,
one first needs a 2-body scattering matrix ( or rather , an `` @xmath1-invariant yang - baxter operator '' , see sect .
[ subsection : invariant - yb - ops ] ) satisfying suitable properties such as yang - baxter - relation , crossing symmetry , analyticity , unitarity , etc . in turns out that the precise algebraic and analytic properties required to make the method work are related to each other in a rather intricate way , and one does not , a priori , see an obvious way to generate simple solutions to the requirements .
one result of our paper is to provide such a yang - baxter operator for the spin-@xmath20 principal , complementary and discrete series representations of @xmath1 in sect .
[ section : ybops - lorentzgroup ] .
this is facilitated by using an invariant geometrical description of the corresponding representations due to bros , epstein , and moschella @xcite ( sect .
[ section : uir - lorentzgroup ] ) . our yang - baxter operator @xmath0 can hence be used to define left- and right half - local operator algebras for this model , as we show in a general framework in sect .
[ section : crossing ] , and as concrete models in sect .
[ subsection : chiral - models ] .
if it could be shown that suitable intersections of such algebras are sufficiently large , then this would indeed correspond to ( a chiral half of ) a local 1 + 1 dimensional field theory describing , in some sense , the quantized left ( or right- ) movers of a non - linear sigma model with non - compact target space .
the group @xmath1 is not just the isometry group of @xmath21-dimensional desitter spacetime @xmath18 , but also the conformal isometry group of @xmath22-dimensional _ euclidean _ flat space @xmath3 .
this dual role becomes geometrically manifest if one attaches a pair of conformal boundaries @xmath23 to @xmath18 .
each of these boundaries is isometric to a round sphere @xmath24 , which in turn may be viewed as a 1-point conformal compactification of @xmath3 via the stereographic projection .
the induced action of @xmath1 by this chain of identifications provides the action of the conformal group on @xmath3 .
this well - known correspondence is at the core of the so - called `` ds / cft - correspondence / conjecture '' @xcite .
the idea behind this correspondence is that with the pair of infinities @xmath23 there is associated a pair of euclidean conformal field theories .
one way to see this is that unitary representations of @xmath1 do not correspond to unitary representations of @xmath25 via `` analytic continuation '' @xcite . ] acted upon by @xmath1 . on the other hand , with the `` bulk '' @xmath18 ,
there is associated a corresponding `` string - theory '' with internal symmetry group @xmath1 .
the action of the group essentially connects these two theories . inspired by this circle of ideas
, one might be tempted to ask whether one can , in our setup , also naturally construct a euclidean conformal field theory on @xmath3 associated with our sigma models with target space @xmath18 .
we will demonstrate in the course of this article that this is indeed the case . in our approach ,
the core datum is a 2-body scattering matrix / yang - baxter operator @xmath0 . in section [ euclqft ]
we will outline an abstract procedure how to obtain a corresponding euclidean conformal field theory from such an object .
thus , within our framework , there is a sense in which the essentially algebraic quantity @xmath0 can relate a euclidean conformal field theory in @xmath22 dimensions and a kind of `` string theory '' in @xmath21-dimensional desitter target space , and thereby gives a model for the ds / cft correspondence .
that correspondence manifests itself more concretely as follows in our construction .
if @xmath26 is a lightray coordinate , the left - local chiral fields of the lightray cft are given by [ field1 ] _
r^chir.(u , x ) = _ p , p \{e^iup(xp)^--iz^_r(p , p ) + } , where @xmath27 is a point in desitter space and @xmath28 are `` desitter waves '' of `` momentum '' @xmath29 analogous to plane waves in minkowski spacetime .
the creation operators @xmath30 create a `` particle '' of lightray - rapidity @xmath31 and desitter `` momentum '' @xmath29 and obey a generalized zamolodchikov - faddeev algebra , @xmath32 , where @xmath33 is our @xmath0-operator - operator . ] . _
, scaledwidth=95.0% ] the desitter momentum @xmath29 is an element of the projective lightcone @xmath34 .
it can be identified with a point @xmath35 in @xmath22-dimensional euclidean space @xmath36 , as depicted in the figure above .
this `` celestial sphere '' is identified with @xmath37 .
if we restrict @xmath38 to @xmath19 discrete values @xmath39 and set @xmath40 under the correspondence @xmath35 , we can also define a multiplet of @xmath19 _ euclidean _ quantum fields on @xmath3 as [ field2 ] ^eucl._r , j ( ) = z^_r , j ( ) + z_r , j ( ) , j=1 ,
the idea is that the desired correspondence ( duality ) maps the states created by the operators @xmath41 to the states created by the operators @xmath42 in the `` continuum limit '' @xmath43 .
the fields at past null infinity @xmath44 are related by a tcp operator @xmath45 which is introduced in the main text .
the main input into all our constructions is a solution of the yang - baxter equation ( ybe ) with additional symmetries . in this section ,
we consider solutions to the ybe that are compatible with a representation @xmath46 of a group @xmath47 and an associated conjugation @xmath48 . in our subsequent construction of field - theoretic models
, we will be interested in the case where @xmath49 , @xmath46 is an irreducible ( spin-0 , principal , complementary or discrete series ) representation of it , and @xmath48 the corresponding tcp operator . in order to compare with @xmath2 sigma models and related constructions , we introduce in section [ subsection : invariant - yb - ops ] yang - baxter operators and functions in the general context of a unitary representation of an arbitrary group @xmath47 , and discuss some examples .
the relevant aspects of the representation theory of @xmath50 are recalled in section [ section : uir - lorentzgroup ] in a manner suitable for our purposes , and the connection to the klein - gordon equation on de sitter space is recalled in section [ subsection : klein - gordon ] .
these representations are then used in section [ section : ybops - lorentzgroup ] , where examples of invariant yang - baxter operators for the lorentz group are presented . in the following , a _
conjugation _ on a hilbert space means an antiunitary involution , and the letter @xmath51 is reserved for the flip @xmath52 , @xmath53 , on the tensor square of a hilbert space @xmath54 .
when it is necessary to emphasize the space , we will write @xmath55 instead of @xmath51 . for identities on various spaces
, we write @xmath56 and only use more specific notation like @xmath57 or @xmath58 where necessary .
[ definition : invariant - yb - operator ] let @xmath47 be a group , @xmath46 a unitary representation of @xmath47 on a hilbert space @xmath54 , and @xmath48 a conjugation on @xmath54 . 1 .
an _ invariant yang - baxter operator _
( for @xmath59 ) is an operator @xmath60 such that 1 .
@xmath0 is unitary .
2 . @xmath61=0 $ ] for all @xmath62 .
3 . @xmath63=0 $ ] .
4 . @xmath64 as an equation in @xmath65 ( i.e. , with @xmath66 ) .
+ the family of all invariant yang - baxter operators for a given representation @xmath46 and conjugation @xmath48 will be denoted @xmath68 .
an _ invariant yang - baxter function _ ( for @xmath46 , @xmath48 ) is a function @xmath69 such that for almost all @xmath70 , * @xmath71 is unitary . *
@xmath72=0 $ ] for all @xmath62 . *
* @xmath74 as an equation in @xmath65 ( i.e. , with @xmath66 ) . * @xmath75 .
+ the set of all invariant yang - baxter functions will be denoted @xmath76 .
our main interest is in invariant yang - baxter _ operators _ , the yang - baxter functions serve as an auxiliary tool to construct them .
independent of @xmath77 , the four operators @xmath78 are always elements of @xmath68 ; these are the trivial unitaries satisfying the constraints ( r1)(r5 ) .
the structure of @xmath68 and @xmath76 depends heavily on the representation @xmath46 and group @xmath47 , as we will see later in examples .
it is clear from the definition that any ( say , continuous ) function @xmath79 defines an invariant yang - baxter operator @xmath80 , and any operator @xmath81 defines a ( constant ) yang - baxter function @xmath82 .
furthermore , any invariant yang - baxter function defines an invariant yang - baxter operator on an enlarged space .
this is spelled out in the following construction , which we will use later on in the context of our qft models .
[ lemma : ybf->ybo ] let @xmath79 ( for some group @xmath47 , on some hilbert space @xmath54 ) , and consider the enlarged hilbert space @xmath83 , with @xmath47-representation @xmath84 and conjugation @xmath85 . on @xmath86 , define the operator @xmath87 then @xmath88 .
the proof of this lemma amounts to inserting the definitions and is therefore skipped .
see @xcite for similar results . for later use
, we mention that we can also work with a different measure @xmath89 than lebesgue measure @xmath90 in this construction .
for example , we can take a finite number @xmath19 of point measures , located at @xmath91 . in that case ,
@xmath92 , with orthonormal basis @xmath93 , invariant under the conjugation @xmath94 . on vectors
@xmath95 , our invariant yang - baxter operators then take the form @xmath96 this construction can also be seen as an example of the partial spectral disintegration formulas considered in @xcite . before presenting examples
, we recall how an invariant yang - baxter operator @xmath0 gives rise to an @xmath0-symmetric fock space , following @xcite . in this fock space construction
, we consider an invariant yang - baxter operator @xmath0 and call its group representation , conjugation , and hilbert space @xmath97 , @xmath98 , and @xmath99 , as these data enter on the one particle level .
as is well known , solutions of the yang - baxter equation ( r4 ) induce representations of the braid group of @xmath8 strands on @xmath100 , by representing the elementary braid @xmath101 , @xmath102 , by @xmath103 .
this representation factors through the permutation group because of ( r5 ) , i.e. we have a representation @xmath104 of the symmetric group @xmath105 on @xmath8 letters on @xmath100 . since @xmath0 is unitary ( r1 ) ,
so are the representations @xmath104 .
we denote by @xmath106 the subspace on which @xmath107 acts trivially , i.e. @xmath108 in view of ( r2 ) , the representation @xmath109 of @xmath47 on @xmath100 commutes with the projection @xmath110 , and hence restricts to @xmath111 .
we denote this restriction by @xmath112 .
furthermore , we define a conjugation @xmath113 on @xmath100 by @xmath114 it is clear that @xmath113 is a conjugation on @xmath100 , and thanks to ( r3 ) , it commutes with @xmath110 and thus restricts to @xmath115 @xcite .
we call this restriction @xmath116 . the @xmath0-symmetric fock space over @xmath99 is then defined as @xmath117 we denote depend on @xmath0 . ]
its fock vacuum by @xmath118 , the resulting `` @xmath0-second quantized '' representation of @xmath47 by @xmath119 , and the resulting conjugation by @xmath120 .
@xmath0-symmetric fock spaces generalize the usual bose / fermi fock spaces , which are given by the special cases @xmath121 . for our purposes ,
the @xmath0-symmetric spaces ( for non - trivial @xmath0 ) will be convenient representation spaces for our models .
we next give some examples of invariant yang - baxter operators and functions .
* example 1 : @xmath122 . *
we consider the group @xmath123 in its defining representation @xmath46 on @xmath124 , with complex conjugation in the standard basis as conjugation .
this is a typical finite - dimensional example , which appears in particular in the context of the @xmath2 sigma models .
it is known from classical invariant theory that the @xmath2-invariance constraint ( r2 ) allows only three linearly independent solutions : the identity @xmath56 of @xmath125 , the flip @xmath51 , and a one - dimensional symmetric projection @xmath126 ( * ? ? ?
one can then check that ( r1)(r5 ) together only allow for trivial solutions , i.e. @xmath127 . however , non - trivial yang - baxter _ functions _
@xmath128 do exist .
a prominent example is @xmath129 which satisfies not only ( r1)(r5 ) , but also the analytic properties ( r6 ) , ( r7 ) that will be introduced in section [ section : crossing ] .
this yang - baxter function describes the @xmath2-invariant two - body s - matrix of the @xmath2-sigma model @xcite .
* example 2 : the `` @xmath130 '' group @xmath131 inner symmetries . * as an example of a different nature , we consider `` @xmath132 '' group , i.e. the affine group @xmath133 generated by translations @xmath134 and dilations @xmath135 on the real line @xmath136 . the physical interpretation is to view @xmath136 as a lightray , which describes one chiral component of a massless field theory on two - dimensional minkowski space .
the group @xmath133 has a unique unitary irreducible representation @xmath137 in which the generator of the translations is positive .
we may choose @xmath138 as our representation space , and then have , @xmath139 , @xmath140 where the variable @xmath141 can be thought of as being related to the ( positive ) light like momentum @xmath142 by @xmath143 .
the conjugation @xmath144 extends this representation to also include the reflection @xmath145 on the lightray . in this example , the invariant yang - baxter operators @xmath146 can all be computed , and in contrast to the @xmath2 case , many such operators exist .
the physically interesting ones are given by multiplication operators of the form @xmath147 as in , where @xmath148 is a scalar function satisfying @xmath149 such `` scattering functions '' include for example the two - body s - matrix of the sinh - gordon model , which is @xcite @xmath150 where @xmath151 is a function of the coupling constant . to generalize to a setting with inner symmetries
, we can also , instead of @xmath133 alone , take the direct product @xmath152 of @xmath133 with an arbitrary group @xmath47 , which is thought of as a global gauge group .
we then consider a unitary representation @xmath46 of @xmath47 on an additional hilbert space @xmath54 , and form the direct product representation @xmath153 on @xmath154 , i.e. @xmath155 for later application , we stress that @xmath54 can still be infinite - dimensional , as it is the case for the irreducible representations of @xmath49 . to also have a tcp operator in this extended setting
, we assume that there exists a conjugation @xmath48 on @xmath54 that commutes with @xmath46 ( i.e. , @xmath46 must be a self - conjugate representation ) , and then consider @xmath156 as tcp operator for @xmath157 . under mild regularity assumptions , one can then show that essentially all invariant yang - baxter operators @xmath158 are again of the form @xmath159 .
that is , @xmath160 acts by multiplying with an ( operator - valued ) function @xmath0 , and this function @xmath0 has to exactly satisfy the requirements ( r1)(r5 ) . in particular , @xmath71 commutes with @xmath161 for all @xmath162 , @xmath62 .
this example is therefore quite different from the previous @xmath2-example : many invariant yang - baxter operators exist , and they are essentially all given by invariant yang - baxter functions via .
both examples can be combined by taking the inner symmetry group as @xmath123 , as one would do for describing the @xmath2-models @xcite .
the construction just outlined here can be used to describe the one - particle space of a ( chiral component of ) a massless sigma model with symmetry group @xmath47 . to prepare our construction of such models for @xmath49
, we review some representation theory of this group next . we now turn to the case of central interest for this article , the ( proper , orthochronous ) lorentz group @xmath49 , @xmath164 . in later sections
, this group will appear either as the isometry group of @xmath21-dimensional de sitter space @xmath165 or as the conformal group of @xmath3 . in this section ,
we first give a quick tour dhorizon of some of its representation theory .
readers familiar with this subject can skip to the next section .
our exposition is in the spirit of @xcite , @xcite ( and references therein ) , and we will use the following notation : capital letters @xmath166 etc . denote points in minkowski space @xmath167 .
the dot product of this minkowski spacetime is defined with mostly minuses in this paper , x y = x_0 y_0 - x_1 y_1 - - x_d y_d .
points in @xmath3 are denoted by boldface letters , @xmath168 , and their euclidean norm is written as @xmath169 .
unitary irreducible representations ( uirs ) of @xmath50 are classified by a continuous or discrete parameter corresponding roughly to the `` mass '' in the minkowski context , and a set of spins corresponding to the @xmath170 casimirs of @xmath171 . in this paper
we will only consider the case of zero spin , there is no spin , and our representations exhaust all possibilities , see @xcite . ] and `` principal- '' , `` complementary- '' and `` discrete series '' representations .
there are many unitarily equivalent models for these representations in the literature , see e.g. @xcite .
the most useful description for our purposes is as follows .
first define the future lightcone c_d^+ = \ { p ^d+1 p p = 0 , p_0 > 0 } in @xmath172-dimensional minkowski space .
we think of @xmath173 as a ( redundant ) version of momentum space in the minkowski context . on @xmath173 , consider smooth @xmath174-valued `` wave functions '' @xmath175 which are homogeneous , [ hom ] ( p ) = ^- -i(p ) , where at this stage , @xmath176 is arbitrary . as the fraction
@xmath177 will appear frequently , we introduce the shorthand @xmath178 .
the collection of these wave functions forms a complex vector space which we will call @xmath179 .
a linear algebraic representation of @xmath180 is defined by pullback , [ udef ] v _ ( ) : _ _ , ( p ) : = ( ^-1 p ) . in order for this to define a unitary representation
, we must equip @xmath179 with an invariant ( under @xmath181 ) positive definite inner product .
it turns out that this is possible only for certain values of @xmath182 .
these are , see below . ] : 1 .
( principal series ) @xmath183 .
( complementary series ) @xmath184 .
( discrete series ) @xmath185 .
two complementary or discrete series representations @xmath186 , @xmath187 are inequivalent for @xmath188 , and two principal series representations @xmath186 and @xmath187 are equivalent if and only if @xmath189 .
we now explain what the inner products are in each case . as a preparation , consider first the @xmath21-form @xmath190 and vector field @xmath191 on @xmath173 defined by = , = p_0 +
@xmath190 is the natural integration element on the future lightcone , and @xmath191 the generator of dilations .
both are invariant under any @xmath180 , i.e. @xmath192 .
we then form @xmath193 where @xmath194 is cartan s operator contracting the upper index of the vector @xmath191 into the first index of the @xmath21-form @xmath190 . a key lemma which we use time and again is the following ( * ? ? ?
* lemma 4.1 ) : [ lemma : closedform ] suppose @xmath195 is a homogeneous function on the future lightcone @xmath173 of degree @xmath196
. then @xmath197 is a closed @xmath22-form on @xmath173 , @xmath198 .
using this lemma , we can now describe the inner products .
+ * a ) principal series : * here the degree of homogeneity of the wave functions is @xmath199 with @xmath182 real .
consequently , the product @xmath200 of two smooth wave functions @xmath201 is homogeneous of degree @xmath196 , so the lemma applies .
we choose an `` orbital base '' r0.3 + @xmath202 of @xmath173 ( i.e. a closed manifold intersecting each generatrix of @xmath173 once , see figure on the right , and appendix a for explicit formulas ) , and define a positive definite inner product by [ inn ] ( _ 1 , _
2(p ) . by lemma [ lemma : closedform ] , this definition is independent of the particular choice of @xmath203 , in the sense that , if @xmath204 is homologous to @xmath203 , then the inner product defined with @xmath204 instead of @xmath203 coincides with
. this fact implies at once that the operators @xmath181 are unitary with respect to this inner product .
since the type of argument is used time and again , we explain the details : @xmath205 the first equality sign is the definition . in the second equality , it is used that @xmath206 , and in the third equality , a change of variables @xmath207 was made . in the last step we used stokes theorem , noting that the integrand is a closed form , and that @xmath203 and @xmath208 are homologous .
the hilbert space of the principal series representation is defined as the completion in the inner product of the space @xmath179 , which we denote by the same symbol . + + pointwise complex conjugation does not leave @xmath179 invariant ( unless @xmath209 ) because the degree of homogeneity is complex , and conjugation changes @xmath182 to @xmath210 .
since @xmath211 is real , this implies that @xmath212 is an antiunitary map @xmath213 , intertwining @xmath186 and @xmath214 . to compare @xmath186 and @xmath214 , it is useful to introduce the integral operator [ idef1 ] ( i_)(p ) : = ( 2)^- _ b ( p ) ( p p)^-+ i ( p),_. if @xmath215 , the poles of the gamma function are avoided .
the integrand has a singularity at @xmath216 which is integrable if @xmath217 ( see appendix a for explicit formulas arising from particular choices of @xmath203 ) .
thus @xmath218 is well - defined as it stands in particular for @xmath219 , corresponding to the case of a complementary series representation , to be discussed below . for the principal series representations
, @xmath182 is real , and we may define by replacing @xmath182 with @xmath220 and then taking the limit @xmath221 ( as a distributional boundary value , see for example ( * ? ? ? * chap .
3 ) ) . this adds a delta function term and yields a well - defined operator @xmath218 for @xmath222 .
finally , for @xmath209 , one has to take into account the gamma factors in when performing the limit @xmath223 . in this case , one obtains @xmath224 ( this follows from the delta function relation in appendix b ) .
after these remarks concerning the definition of @xmath218 , note that for @xmath225 , the value of the integral does not depend on our choice of orbital base @xmath203 , because the integrand clearly has homogeneity @xmath196 in @xmath226 , and is thus a closed form by lemma [ lemma : closedform ] .
[ lemma : tpc - principal ] 1 . in the principal series case ( @xmath183 ) ,
@xmath227 is a unitary intertwining @xmath186 and @xmath214 , i.e. [ intw0 ] v_- ( ) i_= i_v_(),so^(d,1 ) .
furthermore , there holds @xmath228 and @xmath229 .
2 . each principal series representation
@xmath186 is selfconjugate : @xmath230 is a conjugation on @xmath179 commuting with @xmath186 .
@xmath231 by using the invariance @xmath232 and the same type of argument as given in eqs .
, @xmath218 is seen to have the intertwining property .
the equality @xmath233 follows by a routine calculation . to show @xmath234 ( for @xmath235 ) , it is useful to choose a convenient parameterization of @xmath203 . using the spherical parameterization
( see appendix a ) , the identity @xmath234 follows by application of the composition relation .
the special case @xmath224 has been explained above already .
@xmath236 it is clear that @xmath237 is an antiunitary operator on @xmath179 .
as complex conjugation @xmath238 satisfies @xmath239 , one also sees that @xmath237 is an involution , i.e. @xmath240 . * b ) complementary series : * here the homogeneity of the wave functions is @xmath241 , with @xmath242 , so @xmath182 is _ imaginary_. in this case , we can not apply the same procedure as in the case of the principal series to form a scalar product , because the product @xmath243 of two smooth wave functions @xmath201 is not homogeneous of degree @xmath196 , and consequently , lemma [ lemma : closedform ] does not apply . to get around this problem , we can use the operator @xmath244 , which is well - defined also for @xmath219 . as in the case of the principal series , also here
the integral operator @xmath218 has the intertwining property @xmath245 for all @xmath180 .
as @xmath182 is imaginary , the intertwining operator @xmath218 ensures that the function @xmath246 on @xmath247 formed from two wave functions @xmath248 is homogeneous of degree @xmath196 , and lemma [ lemma : closedform ] shows that [ idef ] ( _ 1 , _ 2 ) _ : = _ b ( p ) ( i__1)(p ) is again independent of the choice of orbital base @xmath203 . the same argument as that given in eq .
then also yields that the inner product just defined is invariant under the representation @xmath186 . for the complementary series , we therefore take as our inner product .
we have : for @xmath249 , the inner product is positive definite .
since we are free to choose any orbital base in , , we can make a convenient choice .
if we choose the spherical model , @xmath202 ( see appendix a for the different canonical models ) then our lemma reduces to lemma 5.5 of @xcite .
the proof is however more transparent choosing the flat model , where @xmath250 is parametrized by @xmath251 . using this parametrization
, we find ( , ) _ & = c d^d-1 _ 1 d^d-1 _ 2 |_1 - _ 2|^-(d-1)+2i ( _ 2 ) + & = c d^d-1 | |^-2i |()|^2 0 . here
@xmath252 are positive numerical constants . in the second line we used the plancherel theorem and a well - known formula for the fourier transform of @xmath253 ( see e.g. ex .
vii 7.13 of @xcite ) .
taking into account standard properties of the gamma function , we see that the prefactor of the integral is positive if @xmath219 . in the complementary series ,
the degree of homogeneity is real , and thus complex conjugation is a well - defined operation on @xmath179 .
moreover , complex conjugation commutes with @xmath218 for imaginary @xmath182 .
thus , if we define @xmath254 , then @xmath237 is an antiunitary involution on @xmath179 in the case of the complementary series .
[ lemma : tcp - complementary ] each complementary series representation is selfconjugate : @xmath254 is a conjugation commuting with @xmath186.@xmath255 * c ) discrete series : * here the degree of homogeneity of the wave functions is @xmath199 with @xmath256 . for these values , the gamma - factors in the definition of @xmath218 ( see ) , and hence also in the inner product of the complementary series become singular .
thus , one can not , for this reason alone , define an inner product for the discrete series by analytic continuation of .
the way out is to pass from @xmath179 to an @xmath50-invariant subspace of wave functions for which the scalar product _ can _ be defined by analytic continuation .
since the kernel of @xmath218 is , up to divergent gamma - factors , given by @xmath257 for the discrete series , a natural choice for this subspace is the set of @xmath258 such that [ discr ] ( p ) = ^-(d-1)-n ( p ) , _ b ( p ) ( p p)^n ( p ) = 0 , where the first equality just repeats the homogeneity condition for the case @xmath259 .
the second condition is independent of the choice of @xmath203 , and hence indeed @xmath50-invariant . by abuse of notation
, we denote the set of @xmath175 satisfying again by @xmath179 . for such
@xmath175 , analytic continuation of , to @xmath260 is now possible . since the residue of @xmath48 at @xmath261 is @xmath262 , we find @xmath263 using , one again verifies that the definition of @xmath218 remains independent of @xmath203 , and therefore , by the same argument as already invoked several times , that the inner product is invariant under @xmath181 .
since analytic continuation does not usually preserves positivity , it is non - trivial , however , that this inner product is actually positive definite @xcite . for @xmath264 ,
the inner product is positive definite .
since we are free to choose any orbital base in , we can make a convenient choice .
we choose the spherical model , @xmath202 ( see appendix a for the different canonical models ) , where @xmath265 , with @xmath266 .
we have the series , for @xmath267 ( -1)^n+1 ( 1-x)^n ( 1-x ) = _ m > n x^m . for @xmath268 ,
this series is absolutely convergent , including the limit as @xmath269 , and all its coefficients are evidently positive .
we apply this identity to @xmath270 in the inner product . exchanging integration and summation
it follows that ( , ) _ = _ m > n a_m ^2 0 , where @xmath271 is the rank @xmath272 tensor on @xmath273 given by @xmath274 and @xmath275 denotes the norm of such a tensor inherited from the euclidean metric on @xmath273 . the integral form of the triangle inequality gives @xmath276 , so the series is absolutely convergent , meaning that exchanging summation and integration was permissible .
the case @xmath277 can be treated e.g. using the flat model and applying a fourier transform we omit the details . as in the complementary series ,
the degree of homogeneity is real , and thus complex conjugation is a well - defined operation on @xmath179
. moreover , complex conjugation commutes with @xmath218 for imaginary @xmath182 .
thus , if we define @xmath254 , then @xmath237 is an antiunitary involution on @xmath179 in the case of the discrete series .
this finishes our outline of the representations . in conclusion , we mention that the conjugation @xmath237 , [ thdef ] ( _ ) ( p ) = & + & i(0 , ) or i+ _ 0 , can be interpreted as a tcp operator , at least on the superficial level that it is an antiunitary involution and commutes with the representation , as the reflection @xmath278 on @xmath165 commutes with @xmath50 . these properties are clearly still satisfied if we multiply @xmath237 by a phase factor ( as we shall do in later sections ) . with the help of @xmath237
, one can therefore extend @xmath186 to include the reflection @xmath278 .
, we could in fact extend @xmath186 to a ( pseudo - unitary ) representation of the full lorentz group @xmath279 . ] in this short section we recall the relation between the principal and complementary series representations @xmath186 and classical and quantum klein - gordon fields on @xmath21-dimensional desitter spacetime @xmath18 .
for our purposes , this space is best defined as the hyperboloid _
d = \ { x ^d+1 x x = -1 } embedded in an ambient @xmath172-dimensional minkowski spacetime @xmath167 .
the metric of @xmath165 is simply ( minus ) that induced from the ambient minkowski space .
it is manifest from this definition that the group of isometries of @xmath165 is @xmath279 , where group elements act by @xmath281 . on testfunctions @xmath282 ,
orthochronous lorentz transformations act according to @xmath283 , and we choose an antilinear action of the full spacetime reflection @xmath284 by @xmath285 . the precise relationship between these transformations on @xmath165 and the `` momentum space '' representations @xmath186 from the previous section is , as in flat space , via a special choice of `` plane wave mode functions '' .
these mode functions are defined as follows .
let @xmath29 be any vector in @xmath173 , choose any time - like vector such as @xmath286 in the ambient @xmath167 , and define @xmath287^{-\alpha - i\nu } \ .
\end{aligned}\ ] ] adding a small imaginary part removes the phase ambiguity for @xmath288 when @xmath289 becomes negative .
the limit is understood in the sense of a ( distributional ) boundary value .
the difference between the `` @xmath290 '' ( `` positive frequency '' ) and `` @xmath291 '' ( `` negative frequency '' ) mode arising when we cross to the other poincar patch is basically a phase .
the modes @xmath292 are ( distributional ) solutions to the klein - gordon equation in @xmath27 with mass @xmath293 on the _ entire _ desitter manifold , ( + m^2 ) u^_p = 0 .
conversely , if @xmath258 is smooth , then the corresponding `` wave packet '' u_^(x ) = _ b ( p ) ( p)(x p)^--i_is a globally defined , smooth solution to the kg - equation . to make contact with the representations @xmath186 , we define for @xmath282 @xmath294 where @xmath295 is the @xmath279-invariant integration element on @xmath165 ( fourier - helgason transformation ) . with these definitions , we have the following lemma .
[ lemma : fourier - helgason ] let @xmath282 , and @xmath296
. then 1 .
@xmath297 , 2 . for @xmath298
, there holds @xmath299 .
@xmath300 , where @xmath301 is the phase factor @xmath302 @xmath231 the desitter wave is homogeneous of degree @xmath303 in @xmath29 .
@xmath236 follows directly from the invariance of @xmath190 . for @xmath304
, we first note @xmath305 .
for the complementary series , we have @xmath306 which implies the result in this case . for the principal series , we first recall ( * ? ? ?
* lemma 4.1 ) @xmath307 together with , this gives @xmath308 and the claimed result follows .
we briefly indicate how these facts can be used to define a covariant klein - gordon quantum field on @xmath165 : denoting by @xmath309 the canonical ccr operators on the fock space over @xmath179 ( this is the special case of the `` @xmath0-twisted '' fock space of section [ subsection : invariant - yb - ops ] given by taking @xmath0 as the tensor flip ) , we define @xmath310 or in the informal notation explained in more detail below in sect .
[ section : qft ] @xmath311 .
\end{aligned}\ ] ] it then follows immediately that the field @xmath312 is a solution of the klein - gordon equation @xmath313 , which is real , @xmath314 .
it transforms covariantly under the second quantization of @xmath186 , namely @xmath315 .
furthermore , taking as tcp operator @xmath316 , we have @xmath317 , and hence also the tcp symmetry @xmath318 .
we now take the lorentz group @xmath49 in one of the representations @xmath186 from section [ section : uir - lorentzgroup ] , and associated conjugation @xmath237 , and ask for the invariant yang - baxter operators and functions @xmath319 and @xmath320 , analogously to the examples for @xmath123 and @xmath321 considered in section [ subsection : invariant - yb - ops ] .
our construction will be clearest in a slightly more general setting : instead of a single representation , we consider two representations @xmath322 , @xmath323 from either the principal , complementary or discrete series , i.e. @xmath324 , and then construct an operator @xmath325 intertwining @xmath326 with @xmath327 .
this operator @xmath328 will be an integral operator with distributional kernel @xmath329 , [ rdef ] ( r^_1_2^_1_2)(p_1 , p_2 ) : = _ b b ( p_1 ) ( p_2 ) r^_1_2(p_1 , p_2 ; p_1 , p_2 ) ^_1_2(p_1,p_2 ) .
the degrees of homogeneity @xmath330 of the kernel in its four variables @xmath331 will be @xmath332 for @xmath333 in the principal or complementary series , this implies immediately that the integrand of is homogeneous of degree @xmath196 in both @xmath334 and @xmath335 , so that does not depend on the choice of orbital base @xmath203 by lemma [ lemma : closedform ] .
furthermore , it follows that @xmath336 , @xmath337 , lies in @xmath338 , i.e. @xmath328 is a map @xmath339 .
the same conclusions also hold when one of the parameters @xmath333 ( or both of them ) belong to the discrete series . here
one has to check in addition that the constraint is preserved by the integral operator .
this follows from the relation .
the integral kernel will be taken of the form @xmath340 , where the exponents @xmath341 are complex numbers to be determined .
this ansatz presumably does not really imply any serious loss of generality , because a general invariant kernel may be reduced to such expressions via a mellin - transform along the lines of @xcite .
since it only contains lorentz invariant inner products , it follows immediately that the corresponding integral operator intertwines @xmath326 with @xmath327 whenever it is well - defined . imposing the above degrees of homogeneity in @xmath342 onto this kernel fixes the powers @xmath343 up to one free parameter , which we call @xmath344 .
this then gives the integral kernels [ rdef - kernel ] r_(p_1,p_2 ; p_1,p_2 ) = & c__1 , _ 2 ( ) ( p_1 p_2)^-i- i_1 - i_2 ( p_1 p_1)^-+i+ i_1 - i_2 + & ( p_2 p_2)^-+ i - i_1 + i_2 ( p_1 p_2)^-i+ i_1 + i_2 .
as in @xmath218 , there can be singularities whenever the two momenta in an inner product coincide , and the same regularization as discussed earlier is understood also here .
the constant @xmath345 is taken to be @xmath346 and @xmath141 is taken to be real . with these definitions , we have [ theorem : r - ds ] let @xmath347 , @xmath162 be such that the poles in are avoided , and @xmath348 the integral operators defined above .
then , @xmath70 , 1 .
@xmath349 is unitary .
@xmath350 for all @xmath351 .
@xmath352 @xmath353 with @xmath354 the tensor flip .
+ @xmath352 @xmath355 4 . on @xmath356
, there holds @xmath357 5 .
@xmath358 .
in particular , whenever @xmath359 , the function @xmath360 is an invariant yang - baxter function , @xmath361 .
there holds the normalization @xmath362 the proof of this theorem is given in appendix b. if we go to one of the canonical models for the orbital base @xmath203 ( described in appendix a ) , we get concrete formulas for @xmath363 and the kernel @xmath33 . in the case of the flat model ,
our expression for @xmath33 then coincides , up to a phase , with an expression derived previously @xcite ( see also @xcite ) for the case of the principal series representation .
these authors also proved the yang - baxter equation ( r4 ) , and a version of the idempotency relation ( r5 ) .
their formalism for finding @xmath33 is based on a different model for the representations . as explained before ,
the exponents @xmath341 are fixed by homogeneity requirements in the variables @xmath364 up to one remaining free parameter , which is @xmath365 . setting this parameter to zero ( as required for the yang - baxter equation ( r4 ) )
leads to a trivial solution .
we thus conjecture that @xmath319 contains only trivial operators ( as in the @xmath2 case , example 1 ) . there
do however exist many other invariant yang - baxter functions , because there are two operations we may carry out on the above integral operator without violating the properties ( r1)(r5 ) : scaling of @xmath141 and multiplication by suitable @xmath141-dependent scalar factors .
[ proposition : tweak - r ] let @xmath366 be the integral operator defined by the kernel , @xmath367 , and @xmath368 a function satisfying @xmath369 then also @xmath370 satisfies ( r1)(r5 ) . the proof consists in a straightforward check of the conditions ( r1)(r5 ) , and is therefore omitted .
we conjecture that the operators @xmath370 form essentially all solutions of the constraints ( r1)(r5 ) . the multipliers @xmath371 satisfy exactly the requirements on `` scalar '' yang - baxter functions .
in particular , there exist infinitely many functions satisfying the requirements .
the freedom of adjusting @xmath0 by rescaling the argument and multiplying with such a scalar function will be exploited in the next section .
as explained in section [ subsection : invariant - yb - ops ] , an invariant yang - baxter operator @xmath372 ( def . [ definition : invariant - yb - operator ] ) gives rise to an @xmath0-symmetric fock space on which twisted second quantized versions @xmath59 of the representation @xmath97 and the conjugation @xmath98 act .
these `` covariance properties '' are one essential aspect of the yang - baxter operators in our setting .
the other essential aspect are locality properties , which are linked to specific analyticity requirements on @xmath0 .
these analyticity properties , to be described below , have their origin in scattering theory , where they describe the relation between scattering of ( charged ) particles and their antiparticles @xcite . for the following general discussion ,
we first consider the conformal sigma models with some arbitrary inner symmetry group @xmath47 ( `` example 2 '' of section [ subsection : invariant - yb - ops ] ) , and later restrict to @xmath49 .
our exposition is related to @xcite , where a scalar version of such models was presented , and @xcite , where a massive version with finite - dimensional representation @xmath97 was analyzed .
we consider on the one hand the representation @xmath137 of the translation - dilation group @xmath133 of the lightray on @xmath138 , and the conjugation @xmath373 on that space . on the other hand
, we consider an arbitrary group @xmath47 , given in a unitary representation @xmath46 with commuting conjugation @xmath48 on a hilbert space @xmath54 .
our one - particle space is then @xmath374 , with the representation and the conjugation @xmath375 we pick an invariant yang - baxter function @xmath376 as the essential input into the following construction of quantum fields .
these fields will be operators on the @xmath0-symmetric fock space @xmath377 over @xmath99 given by the invariant yang - baxter operator @xmath160 defined by @xmath0 via .
the @xmath0-symmetric fock space carries natural creation / annihilation operators : with the help of the projections @xmath378 , we define as in @xcite , @xmath379 , @xmath380 with these definitions , @xmath381 is an annihilation operator ( in particular , @xmath382 ) , and @xmath383 is a creation operator ( in particular @xmath384 ) .
it directly follows that , @xmath379 , @xmath62 , @xmath385 where @xmath386 denotes either @xmath387 or @xmath388 . to introduce a field operator on the lightray , we define in is motivated by the fact that we want to study a chiral field , which has an infrared singularity at zero momentum because of the divergence in the measure @xmath389 for @xmath390 .
this problem is most easily resolved by passing to the current of this field , which amounts to taking derivatives of test functions .
as these derivatives result in the prefactor @xmath391 , subsequent formulas will be easier if we include this factor from the beginning . ] for a test function @xmath392 @xmath393 in analogy to the fourier - helgason transforms , these functions lie in the representation space @xmath138 , and the definition is covariant under @xmath133 in the following sense : @xmath394 analogously , the tcp transformed function @xmath395 yields @xmath396 , where the minus sign is due to the fact that we consider the current . given any vector @xmath397 , we define the field operators ( cf .
@xcite ) @xmath398 we may think of the vector @xmath399 as a label for the different `` components '' @xmath400 of the field @xmath401 .
note , however , that @xmath54 can be infinite - dimensional , so that @xmath401 can be a field with infinitely many independent components . by proceeding to delta distributions @xmath402 , sharply localized at a point @xmath403 on the lightray , we may also describe this field in terms of the distributions @xmath404 . in the present general setting
, one can show that the field operators transform covariantly under @xmath137 and @xmath46 , but not under the tcp operator @xmath45 .
furthermore , @xmath405 on an appropriate domain .
we do not repeat the calculations from @xcite here ( see , however , section [ section : qft ] for a concrete version of these properties ) , but rather focus on the locality aspects . to begin with , one realizes that @xmath401 is a non - local field unless @xmath406 , in which case it satisfies canonical commutation relation and reduces to a free field . for general @xmath0 ,
the locality properties of @xmath401 are best analyzed by introducing a second `` tcp conjugate '' field , @xmath407 which in its smeared version is , @xmath392 , @xmath408 this field shares many properties with @xmath401 , and in particular , also transforms covariantly under @xmath137 and @xmath46 .
we next want to analyze the commutation relations between @xmath401 and @xmath409 . to this end
, one first computes that the `` tcp conjugate creation operator '' acts on @xmath410 according to @xmath411 i.e. in comparison to the left action of @xmath383 , this operator `` creates from the right '' .
in particular , @xmath412 , @xmath413&=0\,,\qquad [ z_r(\bxi),\,\theta z_r(\bxi')\theta]=0\,.\end{aligned}\ ] ] to control the mixed commutators @xmath414 $ ] , the following definition is essential .
[ definition : crossing - symmetric - yb - operator ] an invariant yang - baxter function @xmath79 ( for some group @xmath47 , on some hilbert space @xmath54 ) is called _ crossing - symmetric _ if it satisfies the following two conditions .
* @xmath415 extends to a bounded analytic function on the strip @xmath416 .
* there holds the crossing symmetry condition , @xmath162 , @xmath417 in case that @xmath418 is finite - dimensional , these requirements coincide with the corresponding ones in ( * ? ? ?
2.1 ) , where the conjugation was taken as @xmath419 , @xmath420 , with the components @xmath421 referring to a fixed basis and @xmath422 an involutive permutation of @xmath423 .
the significance of ( r6)(r7 ) is best explained in terms of tomita - takesaki modular theory ( see , for example , @xcite ) , as we shall do now .
note , however , that in the explicit examples to be considered in section [ section : qft ] , we will also give a purely field - theoretic formulation ( theorem [ theorem : commutation - concrete ] ) .
for the following argument , we adopt the concept of `` modular localization '' @xcite .
the main idea is to anticipate a quantum field theory , defined in terms of a system of local von neumann algebras @xcite , and the connection between the modular data of the algebra corresponding to the half line @xmath424 and the one - parameter group of dilations @xmath425 which leaves @xmath424 invariant , and the reflection @xmath426 , which flips @xmath424 into @xmath427 ( `` bisognano - wichmann property '' , @xcite ) .
we define @xmath428 and view this as either an operator on @xmath138 , or on @xmath429 , where it just acts on the left factor . if @xmath392 is supported on the right , @xmath430 , the function @xmath431 has an @xmath432-bounded analytic continuation to the strip @xmath433 , with @xmath434 , @xmath162 . for @xmath435 supported on the left instead , @xmath436 , the same properties hold for @xmath437 ( * ? ? ?
* lemma 4.1 ) . in terms of the operator @xmath438 ,
this means @xmath439 the conditions ( r6)(r7 ) imply that @xmath0 has matching analyticity properties .
namely , the matrix elements of @xmath440 are , @xmath441 , @xmath442 and therefore analytically continue to @xmath443 in view of ( r6 ) , with the boundary value @xmath444 these relations imply the following theorem .
[ theorem : commutation ] let @xmath445 .
then , in the sense of distributions , @xmath446=0\quad\text{for}\quad u < u'\ , , \end{aligned}\ ] ] on the space of vectors of finite fock particle number .
the proof follows the same strategy as @xcite , generalized in @xcite .
we introduce as a shorthand the vector - valued functions @xmath447 , @xmath448 , with field operators @xmath449 and @xmath450 .
here @xmath451 are scalar testfunctions , and to prove the theorem , we have to demonstrate @xmath452=0 $ ] for @xmath430 , @xmath436 .
we pick @xmath453 and compute using , , and the definitions @xmath454\,{\boldsymbol{\varphi}}\rangle & = \langle{\boldsymbol{\psi}},\,\left([{z^{\dagger}}_r({{\boldsymbol{g}}}^+),\theta z_r(\bof^-)\theta]+[z_r(\theta{{\boldsymbol{g}}}^-),\theta{z^{\dagger}}_r(\theta\bof^+)\theta]\right)\,{\boldsymbol{\varphi}}\rangle \nonumber \\ & = \langle\langle{{\boldsymbol{g}}}^+,{\boldsymbol{\psi}}\rangle{\omega},\theta\langle \bof^-,\theta{\boldsymbol{\varphi}}\rangle{\omega}\rangle -2\langle p_2^r({\boldsymbol{\psi}}{\otimes}\theta\bof^-),p_2^r({{\boldsymbol{g}}}^+{\otimes}{\boldsymbol{\varphi}})\rangle \nonumber \\ & \quad+ 2\langle p_2^r(\theta{{\boldsymbol{g}}}^-{\otimes}{\boldsymbol{\psi}}),p_2^r({\boldsymbol{\varphi}}{\otimes}\bof^+)\rangle -\langle \theta\langle \theta\bof^+,\theta{\boldsymbol{\psi}}\rangle{\omega},\langle \theta{{\boldsymbol{g}}}^-,{\boldsymbol{\varphi}}\rangle{\omega}\rangle \nonumber \\ & = \langle \theta{{\boldsymbol{g}}}^-{\otimes}{\boldsymbol{\psi}},\,\boldsymbol{r}({\boldsymbol{\varphi}}{\otimes}\bof^+)\rangle - \langle{\boldsymbol{\psi}}{\otimes}\theta\bof^-,\,\boldsymbol{r}({{\boldsymbol{g}}}^+{\otimes}{\boldsymbol{\varphi}})\rangle \ , .
\label{eq : mixed - comm } \end{aligned}\ ] ] to show that these terms cancel in case @xmath0 satisfies ( r6 ) and ( r7 ) , we consider the first term in , and insert an identity @xmath455 , @xmath456 , in front of @xmath160 . in view of the invariance ( r2 ) of @xmath160
, we then see that this scalar product equals @xmath457 the vectors in the left and right hand entry of the scalar product are analytic in the strip @xmath443 ( taking into account the antilinearity of the left factor , and the fact that @xmath458 in the right factor ) . as explained above , the same analyticity holds for the operator - valued function @xmath459 . taking into account the boundary values @xmath460 and , we see that coincides with @xmath461 which is identical to the second term in .
thus the matrix elements of @xmath462 $ ] between single particle states vanish . using the same arguments as in @xcite ,
one shows analogously that matrix elements between arbitrary vectors of finite particle number vanish .
we may interpret this commutation theorem by regarding @xmath463 as localized in the left halfline @xmath464 , and @xmath465 as localized in the right halfline @xmath466 .
this interpretation is consistent with both , covariance and locality .
the fields @xmath401 , @xmath409 should however not be regarded as the `` physical '' quantum fields of the model , but rather as auxiliary objects @xcite . to proceed to the physical observables / fields , localized in finite intervals on the lightray , it is helpful to use an operator - algebraic setting . by construction ,
the field operators satisfy ( on a suitable domain ) @xmath467 if @xmath468 .
one can show by the same method as in @xcite that @xmath469 is then essentially selfadjoint . in this case
, we can form the unitaries @xmath470 by the functional calculus , and pass to the generated von neumann algebra @xmath471 to conclude the present general section , we point out a few further properties of the algebra @xmath472 , including haag duality with the algebra generated by the reflected field . 1 .
the fock vacuum @xmath473 is cyclic and separating for @xmath472 .
the modular conjugation of @xmath474 is @xmath475 , and the modular operator is the previously defined @xmath438 .
3 . the commutant of @xmath472 is @xmath476 @xmath231 it follows by standard arguments that @xmath473 is cyclic for @xmath472 and @xmath477 @xcite . as in @xcite , one shows by an analytic vector argument that the unitaries @xmath478 and @xmath479 commute for @xmath430 , @xmath436 .
hence @xmath480 , and @xmath473 is also separating for @xmath472 ( and @xmath481 ) .
@xmath304 is a straightforward consequence of @xmath236 by tomita - takesaki theory and the definitions .
the proof of @xmath236 follows by the same line of arguments as in @xcite . in this section
we construct examples of crossing symmetric yang - baxter functions @xmath482 for the @xmath50 representations and conjugations @xmath186 , @xmath237 from section [ section : ybops - lorentzgroup ] .
this will be done by exploiting the freedom to adjust our basic @xmath50-invariant yang - baxter function @xmath0 by scaling the parameter @xmath141 and multiplying by a suitable function of @xmath141 ( proposition [ proposition : tweak - r ] ) .
we define @xmath483 where @xmath484 as before , and @xmath485 with @xmath486 .
@xmath487 is a real parameter chosen below depending on the representations .
it is seen that the infinite product converges absolutely in each case considered .
it is shown in appendix b that @xmath371 satisfies the requirements of prop .
[ proposition : tweak - r ] . [ theorem : crossing ] let @xmath488 label two principal series representations , @xmath162 , and @xmath489 the integral operators defined above with @xmath490 in
. then * @xmath491 extends to an analytic bounded function on the strip @xmath492 , and * there holds the crossing symmetry , for the boundary values : if @xmath493 , @xmath494 @xmath495 the same holds true if @xmath496 belong to a complementary or discrete series representation , and we set @xmath497 in .
thus for all principal and complementary series representations , @xmath498 is crossing - symmetric in the sense of def .
[ definition : crossing - symmetric - yb - operator ] . in appendix b
, we also provide an alternative integral representation of the factor @xmath499 . in case
we allow for arbitrary combinations of principal , complementary and discrete series representations @xmath333 , the ( rescaled ) gamma factors in can produce poles in the strip @xmath492 , which is the reason for our restriction to two principal series representations , or a single complementary resp .
discrete series representation .
the function @xmath371 is precisely constructed in such a way that the crossing relation ( r7 ) holds .
there remains however a large freedom to modify @xmath371 without violating the conditions ( r1)(r7 ) .
in fact , we may multiply @xmath500 by another function @xmath501 , which is analytic and bounded on the strip @xmath492 , and has the symmetry properties @xmath502 there exist infinitely many of such `` scattering functions '' , they are given by all inner functions @xmath503 of the strip @xmath492 with the symmetry properties @xmath504 . using the canonical factorization of inner functions ( see , for example , @xcite )
, one can then write down explicit formulas for @xmath503 . in particular , if @xmath503 contains no singular part
, it has the form @xmath505 where @xmath506 , and the zeros @xmath507 have to satisfy @xmath508 and certain symmetry and summability conditions to ensure @xmath504 and convergence of the product @xcite .
we now use the crossing symmetric @xmath50-invariant yang - baxter function @xmath509 to build concrete quantum field theoretic models .
a first class of models will be constructed within the general setup of section [ section : crossing ] , where now the group is taken as @xmath49 , and we restrict ourselves to a principal or complementary series representation @xmath186 . in a second section , we show that by a variant of this construction , one also gets euclidean conformal field theories in @xmath22 dimensions . our first family of models is a concrete version of the abstract field operators in section [ section : crossing ] .
the symmetry group is here the direct product @xmath510 of the translation - dilation group ( acting on lightray coordinates ) , and the lorentz group ( acting on de sitter coordinates ) .
group elements will be denoted as @xmath511 in an obvious notation .
this group is represented on @xmath512 by the representation @xmath513 as in section [ section : crossing ] , where @xmath186 may belong to the principal or complementary series .
that is , the single particle vectors are scalar functions @xmath514 depending on two `` momentum coordinates '' , @xmath515 and @xmath516 , and @xmath517 as tcp operator on this space , we take @xmath518 , where @xmath238 denotes complex conjugation , @xmath237 is defined in and @xmath519 is the phase factor .
we then consider the @xmath50-invariant crossing - symmetric yang - baxter function @xmath520 , which defines an @xmath521-invariant yang baxter operator @xmath522 as in section [ subsection : crossing - yb - ops ] . in order not to overburden our notation ,
we have dropped the tilde from @xmath0 , but still mean _ the rescaled and multiplied version from _ . following the general construction ,
we then obtain the @xmath0-symmetric fock space @xmath377 .
an @xmath8-particle vector @xmath523 is here given by a function of @xmath524 momentum space variables , namely @xmath525 , which is square integrable in each @xmath526 and homogeneous of degree @xmath241 in each @xmath527 ( as well as square integrable on any orbital base @xmath203 of the cone @xmath173 ) .
the @xmath0-symmetry is expressed by the equations @xmath528 to be satisfied for each @xmath529 . here , and
in the rest of the section , we will usually use the shorthand @xmath530 for an arbitrary orbital base @xmath203 to simplify the notation .
the representation @xmath46 of @xmath531 takes explicitly the form , @xmath532 , @xmath533 the explicit form of the tcp symmetry @xmath45 is different for the two series : we have @xmath534 with @xmath535 . for the creation / annihilation operators , we write informally , @xmath379 , z^#_r ( ) = _ b d(p ) z^#_r(,p )
^#(,p ) , where , informally , we take @xmath536 , so that the integral informally is independent of the choice of @xmath203 , by the argument based on lemma [ lemma : closedform ] .
the transformation law then takes the form @xmath537 where the `` @xmath290 '' sign is used for the creation operator @xmath388 , and the `` @xmath291 '' sign for the annihilation operator @xmath387 .
the commutation relations are obtained in this informal but efficient notation as [ zccr ] z_r(_1 , p_1 ) z_r(_2 , p_2 ) - r__1-_2 z_r(_2 , p_2 ) z_r(_1 ,
p_1 ) & = & 0 , + z_r^(_1 , p_1 ) z_r^(_2 , p_2 ) - r__1-_2 z_r^(_2 , p_2 ) z_r^(_1 , p_1 ) & = & 0 , with @xmath349 acting as in .
the form of the mixed commutation relation differs according to whether we are in the principal series case ( @xmath183 ) , the complementary ( @xmath249 ) or discrete series case ( @xmath538 ) . in the first case we have z_r(p_1 ,
_ 1 ) z_r^(p_2 , _ 2 ) - r__2-_1 z_r^(p_2 , _ 2 ) z_r(p_1 , _ 1 ) = ( _ 1-_2)(p_1,p_2 ) 1 where @xmath539 is the dirac delta function on @xmath203 ( relative to the integration measure @xmath363 ) when we integrate this identity against smooth test functions on any orbital base @xmath203 . in the case of the complementary series we have instead z_r(p_1 , _ 1 ) z_r^(p_2 , _ 2 ) - r__2-_1 z_r^(p_2 , _ 2 ) z_r(p_1 , _ 1 ) = c_(_1 - _ 2 ) ( p_1 p_2)^-+i 1 .
in the case of the discrete series , we have z_r(p_1 , _ 1 ) z_r^(p_2 , _ 2 ) - r__2-_1 z_r^(p_2 , _ 2 ) z_r(p_1 , _ 1 ) = c_n ( _ 1 - _ 2 ) ( p_1 p_2)^n ( p_1 p_2 ) 1 where @xmath540 .
the differences arise from the differences in the definition of the scalar product in each case , see , , respectively .
all these exchange relations are generalizations of the zamolodchikov faddeev algebra @xcite .
we next describe a concrete version of the quantum field @xmath401 from section [ section : crossing ] , replacing the `` components '' @xmath400 by an additional dependence on a desitter variable .
this is done by replacing the vector @xmath399 by a fourier - helgason transform @xmath541 of some testfunction @xmath282 . in view of lemma [ lemma : fourier - helgason ] @xmath304 , this amounts to the field operator , @xmath392 , @xmath282 , @xmath542 where @xmath543 are defined in , and @xmath544 in . in the following
, it will be convenient to describe the field operator in terms of its distributional kernels , writing @xmath545 .
then we have , informally ( @xmath546 ) , @xmath547 here , the @xmath38-integral is over @xmath136 , and @xmath548 for an arbitrary orbital base @xmath203 , as before .
we collect a few properties of this field in the following proposition .
[ proposition : phir - props ] the field @xmath401 is an operator - valued distribution on @xmath549 with the following properties . 1 .
the field is neutral in the sense that @xmath550 ( on an appropriate domain ) .
2 . the field is @xmath551-covariant , i.e. @xmath552 3 .
the field solves the klein - gordon equation of mass @xmath553 on de sitter space , @xmath554 @xmath231 is evident from , , and @xmath236 is a consequence of .
note that the prefactor @xmath555 is due to the fact that we consider the current .
@xmath304 is satisfied because the de sitter waves @xmath556 are solutions of the klein - gordon equation .
the field @xmath401 is seen to fail the usual condition of einstein causality in both , its lightray and its de sitter coordinate , due to the presence of the @xmath0-factors in .
so in this sense @xmath401 does not , by itself , straightforwardly define a local quantum field neither on the lightray nor on de sitter space .
however , the field satisfies a kind of remnant of the locality condition in the lightray variable . as explained in the abstract setting in section [ section : crossing ] , this is best understood in interplay with its tcp reflected partner field .
the definition translates here to @xmath557 it is clear from this definition that also @xmath409 has the properties @xmath231@xmath304 of proposition [ proposition : phir - props ] .
explicitly , we have from @xmath558 this is different from @xmath559 because @xmath560 , i.e. the zamolodchikov operators do not transform covariantly under @xmath45 . we have the following concrete version of theorem [ theorem : commutation ] . [ theorem : commutation - concrete ] let @xmath561 be arbitrary , and @xmath182 corresponding to a principal or complementary series representation .
then , in the sense of distributions , @xmath562=0\quad\text{for } \;u < u'\ , , \end{aligned}\ ] ] on the space of vectors of finite fock particle number .
this theorem follows from theorem [ theorem : commutation ] as a special case . however , we give an independent , explicit argument which illustrates how the properties of the integral operator @xmath0 enter .
we focus on the principal series for definiteness .
we expand both @xmath563 in terms of the zamolodchikov - faddeev creation / annihilation operators @xmath564 respectively their primed counterparts . the commutator @xmath565 $ ] then gets contributions of the type @xmath566 , [ z_r^\dagger , z_r^{\prime \dagger}]$ ] , as well as @xmath567 , [ z_r , z_r^{\prime \dagger}]$ ] .
it is relatively easy to see that the @xmath568 $ ] and @xmath569 $ ] contributions vanish separately ( see ) .
this is not the case , however , for the remaining mixed contributions , where a non - trivial cancellation between both terms , called `` @xmath290 '' and `` @xmath291 '' , is required .
if we apply these contributions to an @xmath8-particle state @xmath570 , we get a combination of two terms abbreviated as [ 2k ] [ _ r(u , x ) , _r(u,x ) ] _ n = ( ^+ k - ^- k ) _ n . here , each @xmath571 acts as multiplication operator in @xmath572 and as integral kernel ( depending on @xmath573 and @xmath574 ) on @xmath575 .
taking into account the @xmath0-symmetry of the wave functions , the explicit form of those kernels is found to be : @xmath576 for `` @xmath290 '' , whereas for `` @xmath291 '' one has @xmath577 to get to the expression for `` @xmath291 '' , we have also used properties r1 ) and r3 , case 2 ) of the kernel @xmath33 .
a graphical expression for both kernels in the notation of appendix b is given in the following figure .
+ here the dots @xmath578 , @xmath579 , indicate integrations over @xmath203 as explained in appendix @xmath203 , and the two remaining integrals over @xmath580 and @xmath581 are written explicitly .
the dashed lines connecting the @xmath0-kernels to the desitter waves simply indicate that these parts share the same desitter momentum ( @xmath580 and @xmath581 , respectively ) , and the double lined boxes mean the _ full _ integral kernel of @xmath582 respectively @xmath583 , i.e. including all gamma - factors , the crossing function @xmath584 , and the rescaling @xmath585 . in order to see that the contribution from the `` @xmath290''-kernel cancels precisely that from the `` @xmath291 '' kernel
, we now shift the integration contour in @xmath586 upwards to @xmath587 .
the shifted contour lies in the domain of analyticity of each @xmath588 , by r6 ) .
furthermore , under @xmath589 , we have @xmath590 . by assumption @xmath591 , so the real part of this expression is negative and this provides an exponential damping of the integrand for large values of @xmath592 .
the contour shift is thus permissible .
for @xmath593 , crossing symmetry r7 ) ( see ) then implies that @xmath594 can also be expressed as here the lines with @xmath595 denote integral operators @xmath596 . since @xmath597 ( lemma [ lemma : tpc - principal ] ) , the inner lines cancel .
the two outer operators have the effect of switching the sign of @xmath182 in both the desitter waves .
if we now flip the signs on @xmath598 ( taking into account that this changes the @xmath599 prescription on the desitter waves ) , it becomes apparent that the above expression coincides with @xmath600 , so that the two kernels precisely cancel .
this concludes the proof .
the truly local fields / observables of these model are different from the half - local fields @xmath401 , @xmath409 , and can abstractly be described in an operator - algebraic setting .
we therefore proceed from the pair of field operators @xmath401 , @xmath409 to the pair of von neumann algebras @xmath472 , @xmath601 , which by the preceding result are localized in half lines in the lightray coordinate . in order to build from these basic `` half line '' von neumann algebras a net of von neumann algebras @xmath602 , indexed by intervals @xmath603 ,
one has to translate @xmath604 and form intersections .
one defines for the interval @xmath605
@xmath606 then by construction , we have the assignment @xmath602 from open intervals to von neumann algebras forms a @xmath607-covariant local net of von neumann algebras on @xmath377 : 1 . for any open interval @xmath608 , @xmath609 , @xmath610 2 . for two disjoint intervals
@xmath611 , we have @xmath612=\{0\}. \end{aligned}\ ] ] the elements of @xmath613 may be understood as the local fields / observables of these models .
we do not investigate them here , but just mention two important questions in this context : a ) under which conditions are the algebras @xmath613 `` large '' ( for example in the sense that the fock vacuum @xmath473 is cyclic ) ? and b ) : under which conditions does the net @xmath614 extend from the real line to the circle , transforming covariantly even under the 1-dimensional conformal group @xmath615 , acting by fractional transformation @xmath616 with @xmath617 ?
following the same arguments as in the scalar case @xcite ( which build on @xcite ) , one can show that i ) the subspace @xmath618 is independent of the considered interval @xmath619 , ii ) on @xmath620 , the representation @xmath46 extends to psl@xmath621 , and iii ) the net @xmath622 extends to a local conformally covariant net on the circle .
a direct characterization of @xmath620 is however difficult in general in particular because the nuclearity criteria @xcite that can be applied to the @xmath2 models @xcite do not apply here because the representation @xmath186 is infinite - dimensional .
we leave the analysis of these questions to a future investigation . here
we present a variant of our construction which leads to euclidean conformal field theories in @xmath22-dimensions .
as before , the construction is based on an @xmath50-invariant yang - baxter function @xmath0 such as .
we do not rely on the crossing property in this section .
to turn @xmath0 into a yang - baxter operator , we use here the amplification discussed after lemma [ lemma : ybf->ybo ] instead of the coupling to the representation space of the lightray .
that is , we pick some @xmath624 and real numbers @xmath625 . for given @xmath182 in the _ complementary series _ representation of @xmath50 ( the conformal group in @xmath22-dimensional euclidean space ) ,
we define the one particle space as _ 1 = ^n _ , and the invariant yang - baxter operator @xmath626 on @xmath627 , referring to an orthonormal basis @xmath628 of @xmath124 . the @xmath0-symmetric fock space @xmath377 is then defined as before .
we now choose the orbital base @xmath629 to be flat ( see appendix a ) , which amounts to parameterizing @xmath516 as [ px ] p = ( ( ||^2 + 1 ) , , ( ||^2 - 1 ) ) in terms of @xmath630 , and results in a one - particle space of the form @xmath631 .
vectors in this space are @xmath19-component functions @xmath632 , @xmath633 , and the scalar product is ( see and appendix a ) @xmath634 where @xmath635 and @xmath636 in the complementary series .
we therefore have @xmath19 pairs of creation / annihilation operators @xmath637 , and the @xmath19-component quantum fields @xmath638 the @xmath639 ( and @xmath640 ) are operator - valued distributions , but now defined on the _ euclidean space @xmath203 instead of the `` momentum space '' @xmath173_. we we again describe these fields in terms of their distributional kernels @xmath641 and @xmath642 at sharp points @xmath643 , i.e. @xmath644 , etc . as a consequence of the scalar product ,
the faddeev - zamolodchikov operators satisfy the relations [ zccr1 ] z^_i(_1 ) z_j(_2 ) - r__i-_j z_j(_2 ) z^_i(_1 ) & = & c__ij|_1 - _ 2|^-2 + z^_i(_1 ) z^_j ( _ 2 ) - r__i-_j z^_j(_2 ) z^_i(_1 ) & = & 0 . by construction ,
the field operators @xmath640 are real , @xmath645 , and transform in the complementary series representation @xmath186 , i.e. v _ ( ) _ r , j ( ) v_()^-1 = j_()^- _ r , j ( ) , where @xmath646 is the usual action of conformal transformations @xmath180 on @xmath647 , and where @xmath648 is the corresponding conformal factor , see appendix a. the scaling dimension of @xmath640 is therefore @xmath636 . by the same arguments as before , the exponentiated fields @xmath649
are then well defined for any @xmath650 and any @xmath651 .
we define corresponding `` euclidean '' von neumann algebras _ r(o ) = \ { e^i_r , j(f ) f c^_,0(o ) , j=1 , ,
n } , for any open , bounded region @xmath652 .
by construction , conformal transformations act geometrically on the net @xmath653 in the sense that @xmath654 .
if we choose our discretized rapidities @xmath655 to be spaced equidistantly and formally take @xmath43 , then the shifts @xmath656 correspond to symmetries of the net @xmath657 . the simplest case of this construction is given when instead of the integral operators @xmath0 , we take the flip @xmath406 on @xmath658 . in that case , the @xmath33 factor drops out of the commutation relation for the @xmath659 s , and one finds that the vacuum correlation functions of the field @xmath660 are of quasi - free form , i.e. @xmath661 where the sum is over all partitions of the set @xmath662 into ordered pairs , @xmath663 is a real constant , and @xmath664 for all @xmath665 is assumed .
these correlation functions correspond to an @xmath19-dimensional multiplet of a generalized bosonic free euclidean field theory .
the correlation functions are not reflection positive @xcite ( and so do not define a cft in @xmath22-dimensional minkowski spacetime satisfying the usual axioms ) apart from the limiting case @xmath666 corresponding to the standard free field .
locality of the field theory is expressed by the fact that the above correlation functions are symmetric under exchanges @xmath667 . at the level of the von neumann algebras ,
this amounts to saying that , if @xmath668 and @xmath669 are disjoint , the corresponding von neumann algebras commute [ local ] [ _ f(o ) , _
f(o ) ] = \{0 } . in this sense , the euclidean quantum field theory defined by the assignment @xmath670 is `` local '' .
this structure is modified if instead of the flip , we use our integral operators @xmath0 . in that case , the fields @xmath640 are not `` local '' in the sense that the correlation functions are no longer symmetric , and consequently does not hold . to obtain a local euclidean theory
, one could consider the algebras _
r(o ) : = _ r(o ) _ r(o) , where @xmath669 denotes the complement of @xmath671 , and where @xmath672 is the commutant of the corresponding von neumann algebra @xmath673 .
it follows directly from the definition that the net @xmath674 is local and transforms covariantly under the conformal group @xmath50 in the sense that @xmath675 .
local operators @xmath676 in the conformal field theory defined by this new net should be thought of , roughly speaking , as elements in the intersection of @xmath677 for arbitrarily small @xmath668 containing @xmath678 , i.e. in a sense o ( ) _ o _
r(o ) correlation functions @xmath679 of such fields would then again be local in the sense of being symmetric in the @xmath680 .
to make such statements precise , one should on the one hand make sure that the size of @xmath677 is sufficiently large , and one should also make precise what is meant by the above intersection , presumably by making a construction along the lines of @xcite .
in this work we have constructed yang - baxter @xmath0-operators for certain unitary representations of @xmath50 .
the properties of these operators , in particular the yang - baxter equation , unitarity , and crossing symmetry make possible two , essentially canonical , constructions : a ) a 1-dimensional `` light ray '' cft , whose internal degrees of freedom transform under the given unitary representation and b ) a euclidean cft in @xmath22 dimensions in which the group @xmath50 acts by conformal transformations .
both a ) and b ) are constructed from one and the same @xmath0-operator ( in a complementary series representation ) .
theory b ) depends on a discretization parameter @xmath19 corresponding to a set of @xmath19 discretized `` rapidities '' , @xmath681 .
the operator algebras in cases a ) and b ) become formally related when @xmath43 .
in fact , the operator algebra a ) is related to certain left- and right local fields on the lightray , and whereas the operator algebra in case b ) to certain fields of the form .
they are built from certain generalized creation / annihilation operators @xmath682 in case a ) and @xmath683 in case b ) .
these operators satisfy a zamolodchikov - faddeev algebra into which the @xmath0-operators enter .
@xmath29 is a desitter `` momentum '' which corresponds to @xmath678 as in fig . 1 resp .
the index @xmath684 corresponds to the @xmath684-th discretized rapidity , @xmath526 .
the rapidity variable @xmath38 can be thought of roughly speaking as `` dual '' to the lightray variable , @xmath403 ( in the sense of fourier transform ) , whereas @xmath678 is dual to @xmath27 ( desitter points ) ( in the sense of fourier - helgason transform ) .
thus , when the spacing between @xmath655 goes to zero , the operator algebras formally coincide , and we think of this isomorphism as a manifestation of a ds / cft - type duality . in line with this interpretation , one is tempted to think of @xmath19 as a `` number of colors '' by analogy with the ads / cft correspondence .
however , we note that the algebra in case b ) does not have a corresponding symmetry such as @xmath2 acting on the index @xmath684 . what is restored in the limit as @xmath43 is merely a symmetry corresponding to @xmath685 at finite @xmath19 .
we finally note that our construction does _ not _ yield an ordinary local quantum field theory on desitter spacetime . indeed , despite the formal similarities between the expressions for a free desitter quantum field of mass @xmath272 ( see eq . ) and our field ( see eq . ) , we note that these are actually quite different . whereas the former is local in the spacetime sense with respect to the causal relationships in desitter spacetime , the latter is not ( it is only `` half - local '' in the lightray variable @xmath403 , which is absent in eq . )
. it would be interesting to see whether a variant of our method can also produce new local quantum field theories in the ordinary sense in desitter spacetime .
we must leave this to a future investigation .
* acknowledgements : * the research ( s.h . ) leading to these results has received funding from the european research council under the european union s seventh framework programme ( fp7/2007 - 2013 ) / erc grant agreement no qc & c 259562 .
s.h . also likes to thank r. kirschner from leipzig for pointing out to him references @xcite and for explanations regarding rll - relations and related matters .
it appears best to perform some of the calculations involving the form @xmath211 and the orbital base @xmath203 of the future lightcone @xmath173 of @xmath167 by using specific choices for @xmath203 .
it is known since the times of kepler and newton that there are three canonical choices , which correspond to a flat , hyperbolic , or spherical geometry for @xmath203 . 1 . * ( flat geometry , @xmath687 ) .
* we realize @xmath203 as the intersection of @xmath173 with some arbitrary but fixed _
null _ plane in @xmath167 .
a parameterization of @xmath203 is in this case given by @xmath688 .
the induced geometry is seen to be flat .
the point - pair invariant and @xmath22-form @xmath363 are given ( @xmath689 ) in this case by = d^d-1 , p p = |- |^2 .
the group of transformations leaving @xmath203 invariant is evidently @xmath690 , the euclidean group . with the choice @xmath250
, we may identify the representation space @xmath179 with a space of square integrable functions on @xmath203 . under this identification ,
the action of @xmath180 on a wave function @xmath691 is given by ( u _ ( ) ) ( ) = j_()^--i ( ) , where @xmath646 denotes the usual action of a conformal group element on @xmath678 , and where @xmath648 is the conformal factor of this transformation , @xmath692 .
2 . * ( spherical geometry , @xmath693 ) .
* we realize @xmath203 as the intersection of @xmath173 with some arbitrary but fixed _ space like _ plane in @xmath167 .
a parameterization of @xmath203 is in this case given by @xmath694 .
the induced geometry is seen to be a round sphere .
the point - pair invariant and @xmath22-form @xmath363 are given ( @xmath689 ) in this case by = d^d-1 p , p p = 1-p p , where we mean the standard integration element of the round sphere .
the group of transformations leaving @xmath203 invariant is evidently @xmath695 , the rotational group .
3 . * ( hyperbolic geometry , @xmath696 ) .
* we realize @xmath203 as the intersection of @xmath173 with some arbitrary but fixed pair of parallel _ timelike _ planes in @xmath167 .
a parameterization of the two disconnected components of @xmath203 is in this case given by @xmath697 .
the induced geometry is seen to be hyperbolic for each connected component corresponding to @xmath698 , respectively .
the point - pair invariant and @xmath22-form @xmath363 are given ( @xmath689 ) in this case by = , p p = 1 + - , where the integration element is that of hyperbolic space .
the group of transformations leaving @xmath203 invariant is evidently @xmath699 . + a ) flat , b ) spherical , and c ) hyperbolic orbital base .
the proofs of these theorems make use of the following identities : 1 .
( * symanzik triality relation * ) : + [ eq : triality ] & _
^d-1 d^d-1 q ( 1-p_1 q)^w_1 ( 1-p_2 q)^w_2 ( 1-p_3 q)^w_3 + & = ( 2)^ ( 1-p_2 p_3)^w_1 ( 1-p_1 p_3)^w_2 ( 1-p_1 p_2)^w_3 , + for complex parameters @xmath700 satisfying @xmath701 , where the dual parameters are defined by @xmath702 . the integral on the left side is defined by analytic continuation in the parameters by the method described e.g. in @xcite .
the proof of the identity follows from formula ( 5.104 ) of @xcite ( identical with formula ( b22 ) of @xcite ) , after a suitable analytical continuation in the parameters @xmath703 in the formula and an application of the residue theorem .
+ the triality relation is best remembered in graphical form as a `` star - triangle relation '' , + ( 0,0 ) ( 90:1 ) node[pos=.7,left]@xmath704 ; ( 0,0 ) ( -150:1 ) node[pos=.7,below]@xmath705 ; ( 0,0 ) ( -30:1 ) node[pos=.8,above]@xmath706 ; ( 0,0 ) circle [ radius=0.06 ] ; ( p1 ) at ( 90:1.5 ) @xmath707 ; ( p2 ) at ( -150:1.5 ) @xmath708 ; ( p3 ) at ( -30:1.5 ) @xmath709 ; ( q ) at ( -90:0.4 ) @xmath126 ; ( =) at ( 0:2 ) @xmath710 ; + ( 90:1 ) ( -30:1 ) node[pos=.7,above]@xmath711 ; ( -30:1 ) ( -150:1 ) node[pos=.5,below]@xmath712 ; ( -150:1 ) ( 90:1 ) node[pos=.3,above]@xmath713 ; ( p1 ) at ( 90:1.5 ) @xmath707 ; ( p2 ) at ( -150:1.5 ) @xmath708 ; ( p3 ) at ( -30:1.5 ) @xmath709 ; + here a line with parameter @xmath714 between two `` momenta '' @xmath715 denotes a `` propagator '' @xmath716 , a dot means integration over that variable wr.t .
@xmath717 , and the product over all lines is understood .
( * delta function relation * ) : [ delta ] anal .
cont._w 0 \ { ( 1-p_1 p_2)^-+w } = c_n ( 2)^ ^n ( p_1 , p_2 ) , where we mean analytic continuation in the sense of distributions from the domain @xmath718 . for a proof and a mathematically precise explanation of this kind of analytic continuation ,
see e.g. @xcite .
the identity can be demonstrated by applying laplacians to the composition relation below .
+ the following simple consequence of the triality relation and the delta function relation will be needed for the discrete series representations ( where @xmath719 and @xmath720 ) : + [ eq : triality1 ] & _
^d-1 d^d-1 q ( 1-p_1 q)^n ( 1-p_2 q)^w ( 1-p_3 q)^-2-n - w + & = c_n ( 2)^ ^n ( p_2,p_3)(1-p_1 p_2)^2+n .
+ on the right side , we mean by @xmath721 the delta function on @xmath722 relative to the standard integration measure , and by @xmath438 the laplacian on @xmath722 .
the actual value of the constant @xmath723 is needed only for @xmath277 , where it is @xmath724 .
( * composition relation * ) : this relation is obtained by taking the limit @xmath725 in the triality relation and using the delta function relation .
one obtains the integral identity [ eq : composition - relation ] + & _
^d-1 d^d-1 q ( 1-p_1 q)^ w_1 ( 1-p_2 q)^w_2 + & = ( 2)^2 ( p_1 , p_2 ) , + where it is assumed that @xmath726 .
the composition relation follows in the limit @xmath727 from the triality relation . *
( r3 , part i ) * the flip operator simply exchanges the variables , so that on the level of integral kernels , we have @xmath729 . by inspection of the kernel and the constant @xmath730 ,
one then sees that the first equation in ( r3 ) holds . *
( r1 ) , ( r3 , part ii ) , ( r5 ) : * for unitarity ( r1 ) and the tcp symmetry ( r3 , part ii ) , we have to distinguish the principal and complementary series , because the scalar products and conjugations are different for the two series . in case @xmath488
both belong to principal series representations , the scalar product is given by , and we therefore have @xmath731 here the second step follows by direct inspection of the kernel , taking into account @xmath488 .
we thus have @xmath732 in this case . in case
both @xmath733 and @xmath734 belong to the complementary series , the scalar product is more complicated , but the conjugations @xmath735 are simply complex conjugations ( lemma [ lemma : tcp - complementary ] ) , so that the tcp symmetry ( r3 , part ii ) amounts to @xmath736 .
this equation holds true because @xmath737 are real for the complementary series .
let us consider the tcp symmetry for the principal series , @xmath488 .
then the conjugations @xmath738 are given by complex conjugation and the integral operators @xmath739 ( lemma [ lemma : tpc - principal ] ) .
inserting the definitions and making use of the graphical notation introduced with the triality relation , one finds that the integral kernel of @xmath740 is given by ( 0,0 ) ( 2,0 ) node[pos=.5,below]@xmath741 ; ( 2,0 ) ( 4,0 ) node[pos=.5,below]@xmath742 ; ( 4,0 ) ( 6,0 ) node[pos=.5,below]@xmath743 ; ( 0,2 ) ( 2,2 ) node[pos=.5,above]@xmath744 ; ( 2,2 ) ( 4,2 ) node[pos=.5,above]@xmath745 ; ( 4,2 ) ( 6,2 ) node[pos=.5,above]@xmath746 ; ( 2,0 )
( 2,2 ) node[pos=.85,yshift=+1.7ex , sloped , left]@xmath747 ; ( 4,0 ) ( 4,2 ) node[pos=.25,yshift=-1.7ex , sloped , right]@xmath748 ; ( 2,0 ) circle [ radius=0.06 ] ; ( 4,0 ) circle [ radius=0.06 ] ; ( 2,2 ) circle [ radius=0.06 ] ; ( 4,2 ) circle [ radius=0.06 ] ; ( p2 ) at ( -.4,0 ) @xmath708 ; ( q2 ) at ( 6.4,0 ) @xmath749 ; ( p1 ) at ( -.4,2 ) @xmath707 ; ( q1 ) at ( 6.4,2 ) @xmath580 ; ( 0,0 ) ( 2,0 ) node[pos=.5,below]@xmath741 ; ( 2,0 ) ( 4,0 ) node[pos=.5,below]@xmath742 ; ( 4,0 )
( 6,0 ) node[pos=.5,below]@xmath743 ; ( 0,2 )
( 4,2 ) node[pos=.5,above]@xmath752 ; ( 2,0 ) ( 4,2 ) node[pos=.5,sloped , above]@xmath753 ; ( 4,2 ) ( 6,2 ) node[pos=.5,above]@xmath746 ; ( 2,0 ) ( 0,2 ) node[pos=.45,sloped , above]@xmath754 ; ( 4,0 ) ( 4,2 ) node[pos=.25,yshift=-1.7ex , sloped , right]@xmath748 ; ( 2,0 ) circle [ radius=0.06 ] ; ( 4,0 ) circle [ radius=0.06 ] ; ( 4,2 ) circle [ radius=0.06 ] ; ( p2 ) at ( -.4,0 ) @xmath708 ; ( q2 ) at ( 6.4,0 ) @xmath749 ; ( p1 ) at ( -.4,2 ) @xmath707 ; ( q1 ) at ( 6.4,2 ) @xmath580 ; ( =) at ( 7,1 ) @xmath710 ; ( = = ) at ( -1,1)@xmath710 ; ( 0,0 )
( 2,0 ) node[pos=.5,below]@xmath741 ; ( 2,0 ) ( 6,0 ) node[pos=.5,below]@xmath755 ; ( 0,2 ) ( 4,2 ) node[pos=.5,above]@xmath752 ; ( 2,0 ) ( 4,2 ) ; ( 4,2 ) ( 6,2 ) node[pos=.5,above]@xmath746 ; ( 2,0 ) ( 0,2 ) node[pos=.45,sloped , above]@xmath754 ; ( 4,2 ) ( 6,0 ) node[pos=.5,yshift=.2ex , sloped , above]@xmath756 ; ( 2,0 ) circle [ radius=0.06 ] ; ( 4,2 ) circle [ radius=0.06 ] ; ( p2 ) at ( -.4,0 ) @xmath708 ; ( q2 ) at ( 6.4,0 ) @xmath749 ; ( p1 ) at ( -.4,2 ) @xmath707 ; ( q1 ) at ( 6.4,2 ) @xmath580 ; ( 0,0 ) ( 6,0 ) node[pos=.5,below]@xmath745 ; ( 0,0 ) ( 0,2 ) node[pos=.5,sloped , above]@xmath748 ; ( 0,2 ) ( 4,2 ) node[pos=.5,sloped , above]@xmath752 ; ( 0,2 ) ( 6,0 ) node[pos=.5,sloped , below]@xmath757 ; ( 4,2 )
( 6,2 ) node[pos=.45,sloped , above]@xmath746 ; ( 4,2 )
( 6,0 ) node[pos=.25,yshift=1.7ex , sloped , right]@xmath756 ; ( 4,2 ) circle [ radius=0.06 ] ; ( p2 ) at ( -.4,0 ) @xmath708 ; ( q2 ) at ( 6.4,0 ) @xmath749 ; ( p1 ) at ( -.4,2 ) @xmath707 ; ( q1 ) at ( 6.4,2 ) @xmath580 ; ( =) at ( 7,1 ) @xmath710 ; ( = = ) at ( -1,1)@xmath710 ; ( 0,0 ) ( 6,0 ) node[pos=.5,below]@xmath745 ; ( 0,0 ) ( 0,2 ) node[pos=.5,sloped , above]@xmath748 ; ( 0,2 ) ( 6,2 ) node[pos=.5,sloped , above]@xmath755 ; ( 0,2 ) ( 6,0 ) ; ( 6,2 ) ( 6,0 ) node[pos=.25,yshift=1.7ex , sloped , right]@xmath747 ; ( p2 ) at ( -.4,0 ) @xmath708 ; ( q2 ) at ( 6.4,0 ) @xmath749 ; ( p1 ) at ( -.4,2 ) @xmath707 ; ( q1 ) at ( 6.4,2 ) @xmath580 ; here the dotted lines result from two propagators with opposite powers that cancel each other .
the first dotted line produces a factor of @xmath758 , and the second dotted line produces a factor of @xmath759 . taking into account these factors ( and the @xmath48-factor from the initial diagram ) , one then realizes that the last step represents the integral kernel of @xmath760 .
this finishes the proof of ( r3 , part ii ) for two principal series representations . for the mixed case ( one principal series representation and one complementary series representation ) ,
the proof is similar .
returning to ( r1 ) for two complementary series representations , one finds that because of the appearance of the integral operators @xmath218 in the scalar product , the adjoint is given by @xmath761 passing to the graphical notation , this kernel is given by the exact same diagram as in the tcp symmetry proof for the principal series , but with the replacements @xmath762 , @xmath763 , @xmath764 , @xmath765 . converting the last diagram in the earlier calculation into an integral kernel
then yields @xmath766 .
thus , as in the case of two principal series representations , we have @xmath732 .
again , the proof for the mixed case is similar .
to finish the proof of ( r1 ) and ( r5 ) , it now remains to show @xmath767 .
this follows ( for all combinations of principal / complementary series representations ) by application of the composition relation . *
( r4 ) * the integral identity underlying the yang - baxter relation is the triality relation .
one first calculates that the left and right hand sides of coincide on arbitrary @xmath768 if and only if the following two integral kernels ( with the graphical notation introduced for the triality relation ) coincide : note that these integral kernels differ from the ones arising from the yang - baxter equation by factors of @xmath769 and gamma functions of the parameters , but the overall factors are the same for both diagrams .
we now use the triality relation to convert these diagrams into a more symmetrical form . beginning with the left diagram , we first convert the interior triangle to a star , and then the three resulting stars into triangles .
this shows that the left diagram above coincides with setting all @xmath182-parameters to one and the same value , the properties ( r1)(r5 ) imply the properties ( r1)(r5 ) of an invariant yang - baxter function for the representation @xmath186 and conjugation @xmath237 . ( the properties here are slightly stronger because of the division of ( r3 ) into two separate equalities in ( r3 ) . )
the case when one or more @xmath182-parameters correspond to a discrete series representation can be reduced to the previous cases by perturbing the corresponding @xmath182-parameters slightly from their discrete values along the real axis .
then the same arguments as given for the complementary series go through , and the desired identities are obtained in the limit where the relevant @xmath182-parameters go to their discrete values .
one has to take care , however , to apply all identities to suitable wave functions , and , for the complementary series variables , use the constraint before taking the limit .
one also has to check that this constraint is actually preserved by the @xmath0-operator .
this follows from .
the integral kernel is entire analytic in @xmath141 for non - coinciding momenta , and for @xmath772 and principal series representations , all singularities are integrable .
thus the matrix elements of @xmath349 are analytic functions on the strip @xmath773 , and moreover bounded in @xmath141 on this domain .
we now explain the reason for the particular form of the factor @xmath371 . inserting , we see that ( r7 ) is equivalent to @xmath774 since @xmath488 belong to principal series representations , the conjugations are @xmath775 , where @xmath238 denotes pointwise complex conjugation . using this and lemma [ lemma : tpc - principal ] @xmath231 , one checks that amounts on the level of integral kernels to , @xmath162 , @xmath776 , @xmath777 in the graphical notation , the right hand side can be transformed with the triality relation into a factor of @xmath778 has been suppressed in all these diagrams . comparing with the analytically continued matrix elements on the left hand side of
, one then finds that holds if @xmath779 in order not to spoil the analyticity of the matrix elements of @xmath328 , we have to choose @xmath371 analytic on the strip @xmath780 .
furthermore , @xmath371 must satisfy the requirements of proposition [ proposition : tweak - r ] , i.e. it must be symmetric in @xmath333 , and @xmath781 for @xmath162 .
we claim that all requirements are satisfied by @xmath782 where @xmath783 is a real sufficiently large parameter that will be chosen later , and @xmath784 to verify this claim , we need to examine the functions @xmath785 and @xmath786 .
one first checks the poles of the gamma functions and sees that @xmath785 is analytic in the strip @xmath780 if @xmath783 is sufficiently large , for instance if we take @xmath787 .
its logarithmic derivative can trivially be expressed in terms of the digamma function @xmath788 as @xmath789 clearly , this function @xmath790 is analytic in @xmath780 as well . by taking into account the asymptotic expansion @xmath791 as @xmath792 in @xmath793 , and going through all terms , one also finds that @xmath794 vanishes quadratically in @xmath795 in the strip @xmath780 for @xmath796 .
thus @xmath786 is well - defined .
furthermore , we have the symmetry properties @xmath797 and @xmath798 , @xmath799 , which imply that @xmath786 is even and real ( for real arguments @xmath142 ) . since @xmath786 also decays fast because it is the fourier transform of a smooth function , we see that @xmath800 is well - defined , odd , and real ( for real @xmath141 ) .
it then follows that @xmath371 satisfies the requirements of prop .
[ proposition : tweak - r ] .
it remains to check that @xmath371 is bounded and analytic in @xmath780 , and that holds .
regarding analyticity , the integrand of @xmath800 is entire in @xmath141 , and we may estimate the growth of the sine function by @xmath801 .
this growing factor is compensated by the falloff of @xmath802 .
clearly @xmath803 , and furthermore @xmath804 decays like @xmath805 as well this latter fact follows from a contour shift in the fourier integral . together with the remaining decay of the digamma functions , this establishes the analyticity of @xmath800 in the strip @xmath780 . by analogous arguments
, one also shows that @xmath806 is bounded in the strip . to verify
, we compute the fourier transform @xmath807 of @xmath800 .
we will use that since @xmath786 is even , we have @xmath808 , and we will also make use of @xmath809 .
this gives @xmath810 and after an inverse fourier transformation , we arrive at @xmath811 using the definitions of @xmath371 and @xmath785 , the desired equality then follows .
we must also check the analyticity and boundedness of the @xmath48 factors in the definition of @xmath328 and @xmath371 .
these follow from the well - known facts that @xmath48 is non - vanishing , has poles at the non - positive integers , and the standard asymptotic formula ( @xmath812 ) |(x+iy)|
~(2)^ |y|^x- e^-|y|/2 .
it follows that the @xmath48 factors are bounded by @xmath813 for large @xmath814 .
the @xmath38-dependence of coming from the exponentials is analytic @xmath38 once we form the matrix elements of @xmath33 , and these exponentials are also clearly bounded in @xmath38 .
if we take matrix elements with smooth wave functions , we get from these factors decay as @xmath815 where @xmath8 is as large as we wish .
thus , all pieces in @xmath33 are analytic in the strip @xmath780 and decay faster than any inverse power @xmath815 for @xmath816 if we take matrix elements with smooth wave functions . to derive the infinite product formula for @xmath817 quoted in is rather lengthy , and we only sketch the main steps .
first , we expand the digamma functions in the definition of @xmath818 using the well - known series ( z ) = -_e + _ n=0^ ( - ) . substituting this series for each of the terms in the expression for @xmath818 , we find that all contributions from euler s constant @xmath819 and from the sums over @xmath820 cancel each other .
we next calculate @xmath821 by performing the integral over @xmath822 separately for each term in the series ( this is admissible , because both the series and the integral are absolutely convergent ) .
the resulting integrals all have the form _ -^dt = - e^-(n+)|p| , where the residue theorem was used , and where @xmath823 stands for the various constants that appear . the sum over @xmath8
can then be easily done with the aid of a geometric series , resulting in the expression [ fint ] f__1 _ 2 ( ) = 4 _ 0^dp [ ( _ 12^+ p ) + ( _ 12 ^ -p ) - e^-cp(1+e^p ) ] . in order to perform this integral ,
we expand out the factors @xmath824 using a geometric series , resulting altogether in a double series indexed by natural numbers @xmath825 .
the integral can be pulled inside this double series and can then be performed fairly easily for each term .
each such term turns out to be a logarithm , so the double series of these logarithms becomes a logarithm of a doubly infinite product .
the end result can be written as @xmath826 the product over @xmath272 can be performed with the aid of the well - known infinite product ( z+1 ) = _ m=1^ ( 1 + ) ^-1 e^z / m , and this results in the formula for @xmath827 quoted in the main text after choosing for @xmath663 the value @xmath787 . in case
we consider two coinciding complementary series representations @xmath828 , we have to take into account the different conjugation and different scalar product .
one then finds that the same analytic properties , and in particular the same functional equation are required for the factor @xmath829 .
the solution is given by the same infinite product formula quoted in the main text , but we now need to choose @xmath830 .
this guarantees in particular absolute convergence of the infinite product , as one may see using standard asymptotic expansions of the gamma function .
@xmath255 l. d. faddeev .
_ quantum completely integrable models in field theory _ , volume 1 of _ mathematical physics reviews _
, 107155 ( 1984 ) .
in novikov , s.p .
( ed . ) : mathematical physics reviews , vol . 1 , 107 - 155 g. lechner . _ algebraic constructive quantum field theory : integrable models and deformation techniques_. in : advances in algebraic quantum field theory , r. brunetti et.al .
( eds . ) , 397449 , springer ( 2015 ) | we propose a model for the ds / cft correspondence .
the model is constructed in terms of a `` yang - baxter operator '' @xmath0 for unitary representations of the desitter group @xmath1 .
this @xmath0-operator is shown to satisfy the yang - baxter equation , unitarity , as well as certain analyticity relations , including in particular a crossing symmetry . with the aid of this operator
we construct : a ) a chiral ( light - ray ) conformal quantum field theory whose _ internal _ degrees of freedom transform under the given unitary representation of @xmath1 . by analogy with the @xmath2 non - linear sigma model
, this chiral cft can be viewed as propagating in a desitter spacetime .
b ) a ( non - unitary ) euclidean conformal quantum field theory on @xmath3 , where @xmath1 now acts by conformal transformations in ( euclidean ) _ spacetime_. these two theories can be viewed as dual to each other if we interpret @xmath3 as conformal infinity of desitter spacetime .
our constructions use semi - local generator fields defined in terms of @xmath0 and abstract methods from operator algebras . |
in a universe dominated by collisionless , cold dark matter ( cdm ) , the central densities of galaxy halos are high , and rise to form cusps with @xmath1 , as @xmath2 @xcite .
in contrast , a large class of alternatives to cdm , here generically called warm dark matter ( wdm ) , are expected to have lower central densities and constant density cores at small radii @xcite . in the majority of wdm models ,
phase space arguments predict that density cores will be more prominent in lower mass halos .
current observations of the smallest rotationally - supported galaxies are somewhat ambiguous .
some galaxies prefer cored halos , some prefer cusps , and others are well fit by either @xcite .
moreover , the effects of non - circular motions and other systematic issues are yet to be sorted out for these systems .
the dwarf spheroidal ( dsph ) satellite galaxies of the milky way provide a potentially superior laboratory for studying the nature of dark matter .
observationally , their proximity ( @xmath3 kpc ) allows kinematic studies of individual stars .
theoretically , their small masses make them ideal candidates for prominent wdm cores .
moreover , the cdm prediction of cuspy central density profiles is robust for these satellite systems because the cusps are stable to any tidal interactions that may have occurred @xcite .
unfortunately , obtaining dsph density profile slopes has been difficult .
the best current constraints on their mass profiles come from analyses of @xmath4 line - of - sight ( los ) velocities ( e.g. * ? ? ?
* ; * ? ? ?
the stellar velocity anisotropy is a major source of degeneracy in los velocity dispersion modeling , and , as we show below , the log - slope of the underlying density profile remains practically unconstrained even if the number of observed stars is increased by a factor of @xmath5 . in this _
, we examine the prospects of constraining the dark matter density profiles of dsphs by combining stellar los velocities with proper motion measurements .
we show that @xmath4 proper motions at @xmath6 km s@xmath7 accuracy can break the relevant degeneracies and determine the dark matter distribution in the vicinity of the dsph stellar core radius , @xmath8 . specifically both the dark halo density and the local log - slope of the dark halo density profile at @xmath9 may be determined with @xmath10 times higher precision than from los velocity dispersion data alone . using two well - motivated examples of a cusp and core halo ,
we show that the local log - slope measurement can distinguish between them at greater than the 3-@xmath11 level .
we discuss our results in the context of nasa s _ sim planetquest _ ( space interferometry mission ) which will have the sensitivity to measure proper motions of stars at this accuracy in multiple dsphs .
astrometry as a means to constrain the central density profiles of dsphs was previously considered by @xcite . using a two - parameter family of models for the dsphs
, these authors show that proper motions will be enough to completely reconstruct the profile if the underlying shape follows their adopted form
. however , cdm halos , and likely their wdm counterparts , are described by a minimum of four unknown parameters : a scale density , a scale radius , an _
asymptotic _ inner slope , and an _
asymptotic _ outer slope . by marginalizing over all of these parameters as well as the velocity anisotropy of the stars ,
we demonstrate that the asymptotic slopes are never well - determined , even with proper motions .
we show that what is best constrained is the log - slope , @xmath12 , at a radius @xmath13 comparable to the king core radius , specifically @xmath14 .
we note that unlike the asymptotic @xmath15 slope , the log - slope at @xmath16 kpc is known from cdm simulations .
this radius corresponds to @xmath17 of the relevant halo virial radius , and is well - resolved in current simulations .
we define the three - dimensional ( spatial , not projected ) components of a star s velocity to be @xmath18 , @xmath19 , and @xmath20 . the velocity component along the line of sight is then @xmath21 , where @xmath22 and @xmath23 is the line - of - sight direction . the components parallel and tangential to the radius vector @xmath24 in the plane of the sky are @xmath25 and @xmath26 , respectively . for each component ,
the velocity dispersion is defined as @xmath27 .
we will assume @xmath28 .
the velocity dispersion for each observed component can be constructed by solving the jeans equation for the three - dimensional stellar radial velocity dispersion profile @xmath29 and integrating along the line of sight .
we note that even in the case of tidally disturbed dwarfs , @xcite have shown that dsph velocity dispersions are well modeled by the jeans equation , as long as unbound , interloper stars are removed with standard procedures .
we derive the three resulting observable velocity dispersions : @xmath30 here @xmath31 is the stellar velocity anisotropy , @xmath32 is the surface density of stars , and @xmath33 is the three - dimensional density of stars .
it is clear from inspection that each component depends on @xmath34 in a different fashion , and therefore can be used together to constrain its value . for @xmath32 and @xmath35 we use a king profile @xcite , which is characterized by a core radius , @xmath36 , and tidal radius , @xmath37 .
we adopt values that describe the surface density of draco : @xmath38 kpc and @xmath39 kpc .
note that the remaining dsphs have similar king concentrations , @xmath40 , with sextans having the largest ratio @xmath41 .
our results do not change significantly as we vary @xmath42 ( equivalent to looking at different dsphs ) . in eqs .
( [ eq : losdispersion ] ) , ( [ eq : rdispersion ] ) and ( [ eq : phidispersion ] ) , the radial stellar velocity dispersion , @xmath43 , depends on the total mass distribution , and thus the parameters describing the dark matter density profile .
we will consider the following general parameterization of the dark matter density profile , @xmath44^{(c - a)/b}}. \label{eq : densityprofile}\ ] ] here , the value of @xmath45 sets the _
asymptotic _ inner slope , and different combinations of @xmath46 and @xmath47 set the transition to the asymptotic outer slope . for the specific choice @xmath48 , we have an nfw profile @xcite .
we denote this as our _ cusp _ case below . for our _ core _ case
we use @xmath49 , corresponding to a burkert profile @xcite .
we take these two models to be representative of the predictions of cdm and wdm models .
the burkert profile for the core case is motivated by the expectation that wdm halos will mimic cdm halos at large radius .
this was seen in the wdm simulations of @xcite .
the burkert choice is also conservative compared to the often - used isothermal core with @xmath50 , which is more divergent in shape from an nfw and would be easier to distinguish observationally .
regardless , our methods are robust to changes in the underlying form of the density profile .
( [ eq : densityprofile ] ) allows considerable flexibility in overall form , and the five _ shape _ parameters ( @xmath51 ) are in many cases degenerate .
however , there are a number of physically relevant quantities that may be derived for any set of the five shape parameters .
the first is the log - slope of the dark matter density profile , defined as @xmath52 . for the density profile in eq .
( [ eq : densityprofile ] ) this is given by @xmath53 $ ] .
other quantities of physical interest are the integrated mass within a given radius , @xmath54 , and the physical density at a given radius , @xmath55 , which are clearly obtained for a degenerate set of shape parameters .
below , we show that while the _ shape _ parameters are not well constrained by dsph velocity data , the _ physical _ quantities of interest at the scale of the stellar core radius , @xmath56 , may be constrained to high precision .
our goal is to estimate the accuracy with which the velocity components of stars in dsphs can be used to probe the underlying dark matter distribution . we will consider a model with six independent parameters : @xmath45 , @xmath46 , @xmath47 , @xmath57 , @xmath58 , and @xmath59 constant
. we will consider generalized @xmath60 forms below . in order to keep the profile shape relatively smooth ( as is expected for dark matter halo profiles ) we restrict the range of @xmath46 and @xmath47 by adding gaussian priors of @xmath61 .
the errors attainable on these parameters will depend on the covariance matrix , which we will approximate by the @xmath62 fisher information matrix @xmath63 @xcite .
the inverse of the fisher matrix , @xmath64 , provides an estimate of the covariance between the parameters , and @xmath65 approximates the error in the estimate on the parameter @xmath66 .
the cramer - rao inequality guarantees that @xmath65 is the minimum possible variance on the @xmath67th parameter for an unbiased estimator . using @xmath64 in place of the true covariance matrix
involves approximating the likelihood function of the parameters as gaussian near its peak , so @xmath64 will be a good approximation to the errors on parameters that are well - constrained .
the fisher matrix also provides information about degeneracies between parameters but obviously should not be trusted for estimates of the error along these degeneracy directions .
lll|ccccc @xmath68 & 1.3 ( 0.32 ) & @xmath69 & 2.0 ( 2.2 ) & 0.91 ( 1.0 ) & 0.28 ( 0.24 ) & 0.21 ( 0.23 ) & 0.16 ( 0.15 ) + @xmath70 [ @xmath71 & 1.3 ( 5.3 ) & @xmath72 & 0.27 ( 0.38 ) & 0.12 ( 0.17 ) & 0.11 ( 0.11 ) & 0.07 ( 0.08 ) & 0.06 ( 0.05 ) + @xmath73 [ @xmath74 kpc@xmath75 & 4.0 ( 2.5 ) & @xmath76 & 1.6 ( 0.73 ) & 0.73 ( 0.33 ) & 0.16 ( 0.16 ) & 0.13 ( 0.12 ) & 0.09 ( 0.09 ) + @xmath77 [ @xmath78 & 22 ( 28 ) & @xmath79 & 1.6 ( 1.7 ) & 0.71 ( 0.76 ) & 0.68 ( 0.93 ) & 0.45 ( 0.62 ) & 0.38 ( 0.52 ) + @xmath34 & 0 ( 0 ) & @xmath80 & 2.1 ( 1.4 ) & 0.97 ( 0.62 ) & 0.16 ( 0.16 ) & 0.15 ( 0.15 ) & 0.10 ( 0.10 ) + @xmath45 & 1 ( 0 ) & @xmath81 & 4.4 ( 4.6 ) & 2.0 ( 2.1 ) & 0.78 ( 0.87 ) & 0.55 ( 0.66 ) & 0.44 ( 0.51 ) + @xmath57 [ @xmath74 kpc@xmath75 & 1.0 ( 3.2 ) & @xmath82 & 20 ( 9.2 ) & 9.1 ( 4.2 ) & 5.9 ( 2.0 ) & 4.0 ( 1.5 ) & 3.3 ( 1.1 ) + @xmath83 [ kpc ] & 2.0 ( 1.5 ) & @xmath84 & 8.3 ( 5.1 ) & 3.8 ( 2.3 ) & 3.3 ( 1.8 ) & 2.2 ( 1.2 ) & 1.8 ( 1.0 ) + we pick large radial bins to compute the velocity dispersions and check that this uncorrelates the different bins .
then the elements of @xmath85 are given by @xmath86 the sum is over @xmath87 radial bins and @xmath88 refers to the three velocity `` methods '' one line - of - sight and two components in the plane of the sky .
the errors on the velocity dispersion are represented by @xmath89 .
we choose bins of equal width in distance , so that there are approximately an equal number of stars in each radial bin . as long as we distribute equal number of stars in each bin ,
the results we present below are insensitive to the binning scheme , except in the limit of very few bins , or in the limit of small numbers of stars per bin . to model the errors on the velocity dispersions , we define @xmath90 ^ 2 \rangle$ ] .
we assume that the errors on the velocity of each star are gaussian and that the theory error ( from the distribution function ) and experimental error are summed in quadrature , @xmath91 ^ 2 \ , , \label{eq : totalerror}\ ] ] where @xmath92 is the number of stars in each bin ( taken here to be constant from bin to bin ) . here
@xmath93 is the dispersion in each bin determined from eqs .
( [ eq : losdispersion])-([eq : phidispersion ] ) , and @xmath94 represents the measurement error in the velocities of stars in that bin . for los velocities , the spectroscopic resolution implies errors @xmath95 , which are negligible compared to the underlying dispersion , @xmath96 .
sim can achieve @xmath97as yr@xmath7 proper motions for faint stars . at 100 kpc
, this translates to an accuracy of @xmath6 km s@xmath7 .
we will take this as a typical measurement error for velocities in the plane of the sky . the derivatives in eq .
( [ eq : fishergaussian ] ) , and thus the errors attainable on any of the parameters , depend on the location in the true set of parameter space . to examine how the errors vary as a function of the true set of parameters , we choose two fiducial models .
both models produce a `` typical '' dsph velocity dispersion profile , which is roughly flat at @xmath98 out to @xmath37 .
this requirement does not completely fix the profile . for our _ cusp _ model ( nfw )
we add the further restriction that @xmath83 and @xmath57 fall within the expected cdm range @xcite . for our _ core _ model ( burkert )
we set @xmath83 and @xmath57 to give a central phase space density @xmath99 , which can be produced in some wdm models @xcite .
the values of @xmath83 and @xmath57 for both models are listed in table [ tab : parameterstable ] as are the implied log - slopes @xmath100 ( cusp ) and @xmath101 ( core ) at the characteristic radius @xmath102 , which is set below at the value @xmath103 kpc .
the log - slope of the density profile varies with radius , and so does the minimal error attainable on it .
we are interested in searching for the radius , @xmath102 , where the log - slope is best constrained .
[ fig : gammastar ] shows the @xmath104 error on @xmath105 computed as a function of radius using 1000 line - of - sight velocities and 200 proper motions for the cusp ( left ) and core ( right ) models . in both cases ,
the error reaches a minimum of @xmath106 at @xmath107 and we explicitly adopt @xmath108 for the rest of the discussion .
we find that the value of @xmath102 and the minimum error depend somewhat on the fiducial model . as shown in fig .
[ fig : gammastar ] , profiles with cusps are slightly better constrained than profiles with cores , and @xmath102 occurs at slightly smaller radii for cusped models .
we also find that as @xmath42 increases , the minimum error decreases and @xmath102 increases . we have chosen a conservatively small ratio here , @xmath109 .
table [ tab : parameterstable ] summarizes our results .
we list 1-@xmath11 errors attainable on several quantities determined by marginalizing over the 6 fit parameters discussed above .
we present errors for five different combinations for the number of los stars and proper motion stars . in all cases ,
the errors on the `` physical '' parameters , @xmath110 , @xmath111 , and @xmath112 , decrease significantly when proper motions are included .
quantities listed without ( with ) parentheses correspond to the cusp ( core ) halo case .
note that the errors on the halo shape parameters @xmath46 and @xmath47 are essentially the same as the priors we set on them of @xmath61 and hence we do not list them in the table .
a quantity that is well constrained without the addition of proper motion information is the mass @xmath113 , which can be determined to @xmath114 ( @xmath115 ) for cusped ( cored ) models with current data .
the mass can be determined to a remarkable @xmath116 accuracy when 200 sim stars are added .
the maximum circular velocity is relatively unconstrained by los data alone , but becomes determined to within @xmath117 for cusped ( cored ) models with @xmath118 proper motions .
more interesting for the nature of dark matter is the log - slope of the density profile at @xmath102 .
even with 1000 los velocities , the log - slope in the cusp case is virtually unconstrained , but with the addition of @xmath118 sim stars the log - slope is well - determined , e.g. @xmath119 for the case of an underlying cusp .
the power of velocity dispersions to constrain the density and mass at @xmath120 is relatively simple to understand by examining the observable velocity dispersion components ( eqs .
[ eq : losdispersion ] , [ eq : rdispersion ] and [ eq : phidispersion ] ) .
these observable quantities depend on the three - dimensional stellar velocity dispersion , which scales as @xmath121 for a power - law stellar distribution @xmath35 and constant @xmath34 .
the majority of stars reside at projected radii @xmath122 , where the stellar distribution is falling rapidly @xmath123 . in this case , for @xmath124 , the los component scales as @xmath125 and is dominated by the mass and density profile at the smallest relevant radii , @xmath126 . for @xmath127 , @xmath128 and @xmath129
is similarly dominated by @xmath130 contributions .
therefore we expect the strongest constraints on the dark matter profile at @xmath120 . for @xmath131 similar arguments
hold but there is a manifest degeneracy between @xmath34 and the log - slope @xmath132 . as discussed above and illustrated in table 1 , this degeneracy can be broken by including both los information ( eq .
[ eq : losdispersion ] ) with tangential information ( eqs . [ eq : rdispersion ] , [ eq : phidispersion ] ) .
with the constraints on @xmath132 , are we able to distinguish cored from cusped models ?
to answer this question , in figure [ fig : contours ] we present @xmath133 and @xmath134 error contours in the two - dimensional plane of @xmath132-@xmath34 .
this figure clearly shows the utility of combining both los and proper motions . due to the degeneracy with @xmath34 ,
a sample of @xmath135 line - of - sight stars is unable to distinguish between our two fiducial models , though the models can be clearly distinguished with the addition of proper motions . in the @xmath34-@xmath132 plane ,
the principal constraining power comes from combining the los and @xmath136 velocity dispersions because the @xmath34 dependence in these two quantities comes with opposite signs ( eqs .
[ eq : losdispersion ] , [ eq : rdispersion ] ) .
finally , we address the fact that @xmath34 could in principle vary significantly with radius .
we have repeated our analysis assuming that @xmath137 , and marginalize over @xmath138 , @xmath139 and @xmath140 .
we find very small changes to the @xmath141 error .
for reasons similar to those given above , the three different velocity dispersion profiles constrain @xmath142 to high accuracy , breaking the @xmath34-@xmath143 degeneracy .
we have shown that the measurement of proper motions for @xmath4 stars in a typical dsph can be combined with current line - of - sight velocity measurements to constrain the dark halo log - slope to @xmath0 and normalization to @xmath144 at about twice the king radius .
this is a factor of @xmath6 better than currently possible with los velocity dispersion data alone .
the results from such observations will provide a very sensitive test of the cdm paradigm and an incisive tool for investigating the microscopic nature of dark matter .
we estimate that with @xmath3 days of observing time over the 5 year lifetime of sim , it will be possible to obtain the proper motions of @xmath4 stars in multiple dsphs @xcite .
we thank s. kazantzidis , s. kulkarni , s. majewski , and r. munoz for useful discussions .
les is supported in part by a gary mccue postdoctoral fellowship through the center for cosmology at the university of california , irvine .
jsb , les , and mk are supported in part by nsf grant ast-0607746 . | we show that measurements of stellar proper motions in dwarf spheroidal galaxies provide a powerful probe of the nature of dark matter . allowing for general dark matter density profiles and stellar velocity anisotropy profiles ,
we show that the log - slope of the dark matter profile at about twice the stellar core ( king ) radius can be measured to within @xmath0 when the proper motions of 200 stars are added to standard line - of - sight velocity dispersion data .
this measurement of the log - slope provides a test of cold and warm dark matter theories at a sensitivity not possible with line - of - sight velocity dispersion measurements alone .
the upcoming sim planetquest will have the sensitivity to obtain the required number of proper motions in milky way dwarf spheroidal galaxies . |
m.aleknovich et .
al have proposed a backtracking model in @xcite which generalizes both the priority model of borodin , nielson and rackoff @xcite , as well as a simple dynamic programming model due to woeginger @xcite , and hence spans a wide spectrum of algorithms .
borodin then shared his insights into these models in @xcite .
the definition of backtracking model is as follows .
let @xmath2 be an arbitrary data domain that contains objects @xmath3 called data items .
let @xmath4 be a set , representing the set of allowable decisions for a data item .
for example , for the knapsack problem , a natural choice for @xmath2 would be the set of all pairs @xmath5 where @xmath6 is a weight and @xmath7 is a profit ; the natural choice for @xmath4 is @xmath8 where 0 is the decision to reject an item and 1 is the decision to accept an item .
a backtracking search / optimization problem @xmath9 is specified by a pair @xmath10 where @xmath11 is the underlying domain , and @xmath12 is a family of objective functions , @xmath13 , where @xmath14 is a set of variables that range over @xmath4 , and @xmath15 is a set of variables that range over @xmath2 . on input @xmath16 ,
the goal is to assign each @xmath17 a value in @xmath4 so as to maximize ( or minimize ) @xmath18 .
a search problem is a special case where @xmath18 outputs either 0 or 1 . for any domain @xmath19 write @xmath20 for the set of all orderings of elements of @xmath19 .
a backtracking algorithm @xmath21 for problem @xmath22 consists of the ordering functions @xmath23 and the choice functions @xmath24 where @xmath25 .
there are three classes of bt algorithms .
* _ fixed _ algorithms : @xmath26 does not depend upon any of its arguments . *
_ adaptive _ algorithms : @xmath26 depends on @xmath27 but not on @xmath28 . *
_ fully adaptive _ algorithms : @xmath26 depends on both @xmath27 and @xmath28 .
the value of @xmath26 specifies the order to consider the remaining items , given that the choices about the first @xmath29 items have been made ; the value of @xmath30 specifies the order to make possible decisions about @xmath31 . for more detailed explanation of bt model , refer to @xcite .
but it seems reasonable to repeat the definition of computation tree here , which serves as a good interpretation of _
assume that @xmath9 is a bt problem and @xmath21 is a bt algorithm for @xmath9 , for any instance @xmath32 we define the _ computation tree _
@xmath33 as an oriented rooted tree in the following recursive way .
* each node @xmath34 of depth @xmath29 in the tree is labelled by a tuple @xmath35 . *
the root node has the empty label .
* for every node @xmath34 of depth @xmath36 with a label @xmath37 , let @xmath38 be the data item in @xmath39 that goes first in the list @xmath40 .
assume that the output @xmath41 has the form @xmath42 , where @xmath43 . if @xmath44 then @xmath34 has no children .
otherwise it has @xmath45 child nodes @xmath46 that go from left to right and have labels + @xmath47 resp . following is an example diagram of computation tree , nodes for data items , edges for choices .
it s easy to see that it represents an adaptive bt algorithm , but not fully adaptive .
clearly the width of a computation tree is a lower bound of the computational complexity of the corresponding bt algorithm , so we will consider the asymptotic lower bounds of the computation tree width in the following text .
to understand the proof shown later , we give the basic idea to prove such lower bounds first .
it s a game between solver and adversary .
initially , the adversary presents to the algorithm(solver ) some finite set of possible input items , @xmath48 , and the partial instance @xmath49 is empty , @xmath50 is the set consisting of the null partial solution .
the game consists of a series of phases . at any phase @xmath51
, there is a set of possible data items @xmath52 , a partial instance @xmath53 and a set @xmath54 of partial solutions for @xmath53 . in phase @xmath51 , @xmath55
, the solver picks any data item @xmath56 , adds @xmath57 to obtain @xmath58 , and chooses a set @xmath54 of partial solutions , each of which must extend a solution in @xmath59 .
the adversary then removes @xmath57 and some further items to obtain @xmath52 .
the strategy of adversary will usually have the following form .
the number of rounds , @xmath60 will be fixed in advance .
the adversary will choose some @xmath61 such that , for many partial solutions @xmath62 of @xmath53 , there is an extension of @xmath53 to an instance @xmath63 so that all valid / optimal / approximately optimal solutions to @xmath64 are extensions of @xmath62 .
we ll call such a partial solution _ indispensable _ , since if @xmath65 , the adversary can set @xmath52 to @xmath66 so that the algorithm will never result in correct solution .
first let us recall the definition of knapsack problem .
knapsack problem : : is defined by : + * input : @xmath60 pairs of non - negative integers , @xmath67 and a positive integer @xmath68 , @xmath69 represents the weight of the @xmath51th item and @xmath70 represents the value of the @xmath51th item .
@xmath68 is the volume of the knapsack .
* output : @xmath71 , such that @xmath72 is maximized with respect to @xmath73 .
simple knapsack problem : : is defined by : + * input : @xmath60 non - negative integers @xmath74 and a positive integer @xmath68 , @xmath69 is the weight and value of the @xmath51th item and @xmath68 is the volume of the knapsack .
* output : @xmath71 , such that @xmath75 is maximized with respect to @xmath73 .
both problems are np complete , we will consider only the simple knapsack problem , and denote a simple knapsack problem with @xmath60 items and volume of @xmath68 with @xmath76 .
m.alekhnovich et .
al @xcite have proved the following theorem : for simple knapsack problem ( n , n ) , the time complexity of any adaptive bt algorithms is at least @xmath77 we will improve this lower bound using the same technique as the previous work , by optimizing related parameters , formally : for simple knapsack problem ( n , n ) , the time complexity of any adaptive bt algorithms is at least @xmath78 where @xmath79 is any small positive number .
_ proof _ : consider positive numbers @xmath80 satisfying @xmath81 , then @xmath82 , so we can choose another positive number @xmath83 that satisfies @xmath84 and @xmath85 .
these parameters will be fixed later to optimize the lower bound .
let @xmath68 be some large integer , our initial set of items are integers in @xmath86 .
solver takes the first @xmath87 items one by one , and following each one , adversary applies the following rules to remove certain items from future consideration : remove all items that are the difference of the sums of two subsets already seen ; also remove all items that complete any subset to exactly @xmath68 ( ie all items with value @xmath88 , where @xmath19 is a subset of the items considered so far ) .
these rules guarantee that at any point , no two subsets will generate the same sum , and that no subset will sum to @xmath68 .
also notice that this eliminates at most @xmath89 numbers .
to know why , the difference of the sums of any two subsets can be represented as a weighted sum of the numbers seen so far , with three possible weights : 1 , 0 and -1 , so the number of distinct differences of any two subsets is at most @xmath89 ; the number of distinct values that complete any subset to exactly @xmath68 is even less ( @xmath90 ) , thus can be omitted here .
so we will never exhaust the range from which solver can pick the next item provided that @xmath91 .
call the set of numbers chosen so far @xmath9 and consider any subset @xmath92 contained in @xmath9 of size @xmath93 .
our goal is to show that @xmath92 is indispensable ; that is , we want to construct a set @xmath94 of size @xmath95 , consisting of numbers in the feasible input with the following properties . 1 . @xmath96 .
2 . @xmath97 does not contain other subsets that sum to @xmath68 .
the above properties indeed imply that @xmath92 is indispensable since obviously there is a unique solution with optimal value @xmath68 and , in order to get it , @xmath92 is the subset that must be chosen among the elements of @xmath9 .
we thus get a lower bound on the width of the computation tree of any adaptive bt , which is the number of subsets of size @xmath93 in @xmath9 ; namely @xmath98 .
below is a diagram of such construction . for any @xmath99 ,
let @xmath100 , @xmath101 , @xmath102 $ ] .
we show that @xmath103 .
note that numbers in @xmath92 are less than @xmath104 , so @xmath105 , the last step follows from @xmath85 and @xmath106 ; and @xmath107 , the last step follows from @xmath84 and @xmath106 . to construct @xmath108 ,
first choose @xmath109 items in @xmath110 .
after each choice , we will remove some items from the remaining by the following rules .
let @xmath19 be the set of items currently in @xmath97 . 1 . for all @xmath111 ,
remove items of value @xmath112 .
2 . for all @xmath113 ,
remove items of value @xmath114 . since @xmath100
, we can always avoid the points that need to be eliminated , and sum to a number @xmath115 so that @xmath116 .
this can be done by iteratively picking numbers bigger / smaller than @xmath57 according to whether they average to below / above @xmath57 currently . to complete we need to pick two more items @xmath117 that sum to @xmath118 , and
so that they are not the difference of sums of any two subsets of the @xmath119 items picked so far .
assume for simplicity that @xmath120 is an integer .
of the @xmath121 pairs @xmath122 , where @xmath123 , at least one pair @xmath124 will suffice .
( note that , since @xmath125 , the smallest number is @xmath126 , and the largest number is @xmath127 , provided that @xmath68 is sufficiently large .
so we will never go beyond the range of @xmath128 . )
now it has @xmath96 , the only thing left is to prove there does not exist another subset which also sums to @xmath68 .
suppose for contradiction @xmath19 is such a subset , namely @xmath129 , following are three cases : 1 .
neither @xmath130 nor @xmath131 belong to @xmath19 , contradictary to the second rule when picking the first @xmath119 items .
both @xmath130 and @xmath131 belong to @xmath19 , by comparing @xmath19 and @xmath132 , we get two subsets of the first @xmath119 items that sum equally , contradictary to the first rule . 3 .
exactly one of @xmath130 and @xmath131 belong to @xmath19 , by comparing @xmath19 and @xmath132 , we find that @xmath130 or @xmath131 is equal to the difference of the sums of some two subsets of the first @xmath119 items , contradiction .
the rest of the proof is about maximizing the lower bound : @xmath98 , subject to @xmath133 and @xmath81 . according to stirling formula ,
the target function can be simplified : @xmath134 by applying logorithm transform , our goal is equivalent to maximize : @xmath135 subject to @xmath133 and @xmath136 .
note that for convenience , we allow @xmath137 here , the case will be handled later .
take partial differential on @xmath138 , which is : @xmath139 so @xmath140 is strictly increasing respect to @xmath138 , since @xmath141 , by setting @xmath138 to @xmath142 , we only need to maximize : @xmath143 whose differential is : @xmath144 the only root of the above function that lies in @xmath145 is @xmath146 , which maximizes @xmath147 , thus @xmath148 .
so @xmath138 can be arbitrarily close to @xmath149 , and the maximum value of @xmath150 is @xmath151 by substitution , where @xmath79 is any small positive number .
consequently , the optimal lower bound @xmath98 is : @xmath152
backtracking is an important algorithmic scheme which is pervasive in solving hard problems such as propositional satisfiability problem ( sat ) and constraint satisfaction problem ( csp ) . proving lower bounds for these problems under bt model
will both deepen our understanding of the structure and properties of these hard problems and guide our designing of algorithms to solve these problems .
an exponential time lower bound for sat under fully adaptive bt model has already appeared in @xcite .
recently , xu and li have proven an exponential lower bound for a class of random csps ( model rb ) under tree - like resolution @xcite .
since tree - like resolution has a close tie with dpll , a famous backtracking search strategy for sat with corresponding exponential time lower bound in @xcite , it is natural to investigate the time complexity of some class of random csps ( rb model ) under bt model , and we expect that this future work may produce the first exponential time lower bound for some class of random csps under bt model .
we thank professor ke xu for joining our discussion and providing helpful comments .
a.borodin , a.magen , j.buresh-oppenheim , m.alekhnovich , r.impagliazzo and t.pitassi : toward a model for backtracking and dynamic programming , proceedings of 20th annual ieee conference on computational complexity ( 2005 ) 308 - 322 xin li , tian liu , han peng , liyan qian , hongtao sun , jin xu , ke xu and jiaqi zhu : improved exponential time lower bound of knapsack problem under bt model , proceedings of fourth annual conference on theory and applications of models of computation ( 2007 ) 624 - 631 p.pudlak : proofs as games .
american math .
monthly , 23 ( 2000 ) 541 - 550 d.itsykson , e.hirsch and m.alekhnovich : exponential lower bounds for the running time of dpll algorithms on satisfiable formulas , proceedings of 31st international colloquium on automata , languages and programming ( 2004 ) 84 - 96 ke xu , wei li : many hard examples in exact phase transitions , theoretical computer science , 355 ( 2006 ) 291 - 302 a.borodin , c.rackoff and m.nielson : ( incremental ) priority algorithms , algorithmica , 37 ( 2003 ) 295 - 326 g.woeginger : when does a dynamic programming formulation guarantee the existence of a fully polynomial time approximation scheme ( fptas ) ?
informs journal on computing , 12 ( 200 ) 57 - 75 a.borodin : further reflections on a theory for basic algorithms , proceedings of second annual conference on algorithmic aspects in information and management ( 2006 ) 1 - 9 | m.aleknovich et al . have recently proposed a model of algorithms , called bt model , which generalizes both the priority model of borodin , nielson and rackoff , as well as a simple dynamic programming model by woeginger .
bt model can be further divided into three kinds of fixed , adaptive and fully adaptive ones .
they have proved exponential time lower bounds of exact and approximation algorithms under adaptive bt model for knapsack problem .
their exact lower bound is @xmath0 , in this paper , we slightly improve the exact lower bound to about @xmath1 , by the same technique , with related parameters optimized . |
let @xmath1 be the potential energy function of a classical model for a molecular system .
a fundamental challenge is to sample the configurational gibbs - boltzmann ( canonical ) distribution with density @xmath2 where @xmath3 where @xmath4 is boltzmann s constant , @xmath5 is temperature , and @xmath6 is a normalization constant so that @xmath7 has unit integral over the entire configuration space .
a wide variety of methods are available to calculate averages with respect to @xmath7 ; among these , some of the most popular are based on brownian dynamics or langevin dynamics ( defined in the phase space of positions and momenta ) @xcite .
( in this article we focus exclusively on molecular dynamics techniques ; molecular models can also be sampled using monte - carlo methods , and , more generally , using hybrid algorithms which combine molecular dynamics with a metropolis - hastings test in order to correct averages .
for a recent review of such schemes see @xcite . )
recall that brownian dynamics ( overdamped langevin dynamics ) is a system of it - type stochastic differential equations of the form @xmath8 where @xmath9 is the infinitesimal increment of a vector of stochastic wiener processes @xmath10 , and @xmath11 is a positive ( we assume here diagonal ) mass matrix .
a simple and popular method for numerical solution of eq .
[ bd ] is the euler - maruyama method @xmath12 where @xmath13 is a vector of random variables with standard normal distribution .
this produces a sequence of points @xmath14 , which , following a certain relaxation period , are approximately distributed according to the canonical invariant distribution .
euler - maruyama has the property that the time averages along discrete trajectories , in the limit of large time ( under appropriate conditions on the potential @xmath15 and assuming no effects from floating point rounding error ) , have error proportional to @xmath16 .
one of the observations of this article is that the simple modification @xmath17 provides a second order approximation of stationary averages .
we arrive at this scheme by considering the large friction limit in a particular numerical method for langevin dynamics .
recall that langevin dynamics is a stochastic - dynamical system involving both positions and momenta @xmath18 of the form @xmath19{\rm d}t + \sigma m^{1/2}{\rm d}w , \label{lang}\ ] ] where @xmath20 is again a vector of @xmath0 independent wiener processes , and @xmath21 is a free parameter , the friction coefficient .
the methods that are in fact the primary focus of this paper are splitting integrators that decompose the stochastic vector field of langevin dynamics into simpler vector fields which can be solved exactly .
the composition method that results can not be directly related to a stochastic differential equation , so the analogy with backward error analysis for deterministic problems @xcite is incomplete , but nonetheless the invariant measure associated to the numerical method can be derived , using the baker - campbell - hausdorff ( bch ) expansion , as an asymptotic series in two parameters : the stepsize and the reciprocal of the friction coefficient .
the superconvergence result alluded to above is then obtained in the high friction limit .
the langevin stepsize must be understood to be proportional to the square root of the stepsize that appears in eq .
[ bd - mod ] , so in langevin dynamics an effective 4th order approximation is obtained , but only for the marginal configurational invariant distribution , eq .
[ configdist ] .
our approach also provides a simple method for comparative assessment of the invariant measure of a class of langevin integrators .
molecular dynamics is a large family of modelling techniques which is widely used in different application areas and for different purposes @xcite .
this article is addressed specifically to the topic of calculating averages with respect to an invariant distribution , and will probably be of highest interest for applications in molecular sampling , in particular the calculation of averages with respect to the configurational density eq .
[ configdist ] .
the high friction limit renders langevin dynamics unsuitable for dynamical modelling ( except as a method for generating starting configurations for dynamical exploration ) .
it is worth noting that invariant measure computations arise frequently in applications other than molecular modelling , and the techniques described here would be of potential use in many of these . in large scale simulations
, it should be understood that the statistical error ( dependent on the number of samples used ) is typically the dominant concern .
our approach focuses on the truncation error of the invariant distribution , thus the greatest benefit would be seen only when the statistical error is well controlled .
nonetheless we observe in our numerical experiments that the least biased scheme from the point of view of the error introduced in configurational sampling is also as efficient as the alternatives and the most robust with respect to variation of the parameters ( stepsize , friction coefficient ) , thus there is effectively no price for the improvement in accuracy .
moreover , it would be possible to complement the methods proposed here by procedures such as importance sampling @xcite to further reduce the statistical error . with regard to the approximation of canonical averages ,
methods have previously been constructed for brownian dynamics with order @xmath22 and for langevin dynamics with order @xmath23 @xcite , but these require multiple evaluations of the force ; for this reason they are not normally viewed as competitive alternatives for molecular sampling @xcite .
by contrast all of the methods described in this article use a single force evaluation at each timestep .
the approach used here may be compared to other recent works on stochastic numerical methods , and , in particular @xcite .
our technique differs from these in ( i ) the direct focus on the stationary configurational distribution , and ( ii ) the use of the bch expansion .
other articles ( see e.g. @xcite ) which address the invariant measure in langevin - type stochastic differential equations do not use the backward error analysis ( and do not find the superconvergent scheme ) .
the rest of this article is as follows . in section 2
we review necessary background on stochastic differential equations for sampling from the gibbs measure .
section 3 presents our expansion of the associated perturbed invariant measure and calculations involving the use of the baker - campbell - hausdorff theorem .
section 4 describes the reduction of the methods in the case of overdamped langevin dynamics .
section 5 demonstrates the theory obtained using numerical experiments to verify the results .
the ornstein - uhlenbeck ( ou ) stochastic differential equation is an it equation of the form @xmath24 where @xmath25 is a random variable defined for each time @xmath26 , @xmath27 and @xmath28 are positive parameters and @xmath29 represents the infinitesimal increment of a wiener process .
langevin dynamics ( eq . [ lang ] ) combines the ornstein - uhlenbeck stochastic vector field with conservative dynamics . for stochastic differential equations
, the density evolves according to an evolution equation ( the fokker - planck or forward kolmogorov equation ) of the form @xmath30 where @xmath31 is a second order differential operator . in the case of langevin dynamics
, the relevant kolmogorov operator is defined by its action on a function @xmath32 of the variables of the system by @xmath33 where @xmath34 is the mass - weighted laplacian in the momenta : @xmath35 . by choosing @xmath36 one
easily checks that the gibbs distribution with density @xmath37 is a steady state of the kolmogorov equation , where @xmath38 is a normalising constant and @xmath39 represents the hamiltonian energy function . for later reference
, we define averages with respect to the gibbs distribution by @xmath40 assuming @xmath41 is @xmath42 it is possible to demonstrate that the operator @xmath43 is hypoelliptic by using hrmander s criterion @xcite based on iterated commutators , and this implies that the gibbs measure is the unique steady state ( up to normalization ) .
even stronger is the result of @xcite which demonstrates the existence of a spectral gap for the operator @xmath44 .
many of the challenges related to obtaining formal analytical results for stochastic differential equations relate to singularities of the potential and/or the assumption of an unbounded solution domain .
however , with periodic boundary conditions and strong repulsive potentials ( e.g. lennard - jones potentials ) we observe that configurations typically evolve in a bounded set and remain far from singular points ( the radial distribution vanishes in a large interval around the origin ) .
indeed it is a simple calculation to demonstrate that for a lennard - jones system with potentials @xmath45 the expected number of samples required to observe @xmath46 at unit temperature would involve a simulation of duration far greater than the age of the universe , due to the steepness of the potential close to the origin . in our simulations of a small lennard - jones cluster in section 5 , we did not observe a separation of two atoms beneath @xmath47 in all of the nearly @xmath48 timesteps performed to gather statistics ; a separation less than @xmath49 could only be seen at very large stepsizes , when instabilities due to other components of model , e.g. harmonic bonds , would have anyway rendered a typical molecular simulation useless .
thus it is somewhat of a moot point whether we simply assume that configurations stay well away from singular points ( domain restriction ) or that the potential has been smoothly cut off to remove the singularity ; in no case will we encounter the atomic collision singularity in a simulation of the type envisioned here . for the purpose of deriving practical methods
we assume that ( i ) the positions are confined to a periodic simulation box @xmath50 , where is the one - dimensional torus , and ( ii ) @xmath41 is @xmath42 on @xmath51 .
these assumptions , which are realistic in most molecular dynamics applications for the reasons mentioned above , allow us to use recent results in hypocoercivity @xcite to establish the regularity properties of @xmath44 .
specifically , we have as a consequence of these articles , that * @xmath52 has a unique solution in @xmath53 , i.e. , the gibbs density @xmath54 , * @xmath43 has a compact resolvent @xcite , and the gibbs state is therefore exponentially attracting .
note that , if , for some given @xmath55 , there are two solutions of @xmath56 then , using the linearity of the operator these may differ only in a constant scaling of the gibbs density . in a splitting method for a deterministic system @xmath57 ,
one divides the vector field @xmath58 into exactly solvable parts , i.e. @xmath59 , which are treated sequentially within a timestep .
an example of such a splitting method for the hamiltonian system with energy @xmath60 is the `` symplectic euler '' method defined by @xmath61 , @xmath62 . by dividing the vector field as @xmath63 , solving each vector field in turn
, we obtain the so - called position verlet method , and by switching the roles of @xmath64 and @xmath65 we obtain the velocity verlet method .
splitting methods like these are explicit and this feature is of particular importance in molecular dynamics , where the force calculation is the usual measure of per - timestep computational complexity . in a similar way
, langevin dynamics may be treated by splitting @xcite .
for example , one may divide the langevin system ( eq .
[ lang ] ) into three parts @xmath66= \underbrace{\left [ \!\!\begin{array}{c } m^{-1 } p \\0\end{array}\!\
! \right ] { \rm d}t}_{\rm a } + \underbrace{\left [ \!\ ! \begin{array}{c } 0 \\-\nabla u\end{array}\!\ ! \right ] { \rm d}t}_{\rm b } + \underbrace{\left [ \!\ ! \begin{array}{c } 0\\ -\gamma
p{\rm d}t + \sigma m^{1/2}{\rm d } w \end{array } \!\!\right ] , } _ { \rm o } \label{gla2}\ ] ] and each of the three parts may be solved ` exactly ' . in the case of the ou part ( labelled here simply as o )
we mean by this that we realize the stochastic process by the equivalent formula @xmath67 where @xmath68 is a vector of uncorrelated independent standard normal random processes ( white noise ) .
one method based on the splitting in eq .
[ gla2 ] is defined by the composition @xmath69 where @xmath70 represents the phase space propagator associated to the ( deterministic or stochastic ) vector field @xmath58 , and we use @xmath71 as the timestep in langevin dynamics ( later we use @xmath16 for the timestep of an associated brownian dynamics ) .
the deterministic part is approximated by the position verlet method .
this is referred to as a geometric langevin algorithm of order two ( gla-2 ) following @xcite ; an alternative is to use velocity verlet for the hamiltonian part ( @xmath72 ) .
a simple generalization of gla methods is obtained by interspersing integrators associated to parts of the hamiltonian vector field with exact ou solves , thus we have @xmath73 and @xmath74 defined in an analogous way .
recent work in @xcite similarly uses exact ou solves to give an integrator equivalent to @xmath75 ( in our notation ) , though with a reparameterised timestep .
the analysis technique we use can employed to study many such integrators , though for brevity we will limit this article to discussion only on a select few interesting cases .
an alternative integrator termed the stochastic position verlet ( spv ) method @xcite , relies on the splitting @xmath76 = \left [ \begin{array}{c } m^{-1 } p \\0\end{array } \right ] { \rm d}t + \left [ \begin{array}{c } 0 \\-\nabla u{\rm d}t -\gamma p{\rm d}t + \sigma m^{1/2 } { \rm d } w \end{array } \right ] .\ ] ] spv is not a ( generalized ) gla - type method , although it , as each of the generalized gla schemes , is _
quasisymplectic _ in the language of @xcite .
likewise the commonly used method of brunger , brooks and karplus ( bbk ) @xcite is not of the ( generalized ) gla family .
details of all methods examined are given in the appendix . to slightly simplify the presentation that follows
, we make the change of variables @xmath77 , @xmath78 , with a corresponding adjustment of the potential ; this is equivalent to assuming @xmath79 .
we shall work here with formal series expansions , however we expect that our derivation could be rigorously founded using techniques found in @xcite and @xcite
. associated to any given splitting - based method for langevin dynamics , we define the operator @xmath80 that characterises the propagation of density by an expansion of the form : @xmath81 for example , for the method labelled by the string abao , we have @xmath82 the perturbation series may be found by successive applications of the bch expansion @xcite and linearity properties of the kolmogorov operator .
however , unlike in the deterministic case , the terms that appear in the series can not be associated to modified vector fields or even sdes @xcite .
note that , when iterated @xmath83 times , the method abao produces a sequence of the form @xmath84^n
e^{\frac{\delta t}{2 } { \cal l}_{\rm b}}e^{\frac{\delta t}{2 } { \cal l}_{\rm a } } e^ { { \delta t } { \cal l}_{\rm o } } \ ] ] thus , with a minor coordinate transformation , the dynamics sample the same invariant density as baoab .
similarly babo and obab are essentially the same method as aboba . for this reason we concentrate in the remainder of this article on aboba and baoab .
for these two methods , the symmetry implies that the odd order terms in eq .
[ opexp ] vanish identically using the jacobi identity in the bch expansions . after deriving @xmath80 in this way
, we seek the invariant distribution which satisfies @xmath85 . for the baoab and aboba methods
, we make the ansatz that the invariant measure of the numerical method has the simple form @xmath86).\ ] ] although some technical issues might be encountered , we believe that the existence of such an expansion can be made rigorous using techniques found in @xcite , based on the regularity of the operator @xmath87
. we may rewrite this as @xmath88 this means that the equation @xmath89 becomes @xmath90 equating second order terms in @xmath71 gives @xmath91 the equation @xmath92 has unique solution @xmath93 , up to a constant multiple .
hence the homogeneous solution to the above pde is @xmath94 , for some constant @xmath95 ; we therefore require a particular solution @xmath65 of eq .
[ eqn::pde ] . according to the fredholm alternative
, the equation has a solution provided that , for any solution of @xmath96 we have @xmath97 as the only solutions of @xmath98 are the constants , we require @xmath99 for a symmetric splitting method such as the baoab method , recall that we can use the baker - campbell - hausdorff @xcite formula to find the overall one - step perturbation operator for the scheme . for linear operators @xmath100 , @xmath101 and @xmath6 , we have the relation @xmath102 where @xmath103 \right ] \right .
+ \left[y , \left [ y , x \right ] \right ] + \left[z , \left [ y , x \right ] \right]+ \left[y , \left [ z , x \right ] \right ] \right .
\\ & \left .
- \tfrac{1}{2 } \left[y , \left [ y , z \right ] \right ] - \tfrac{1}{2 } \left[x , \left [ x , z \right ] \right ] - \tfrac{1}{2 } \left[x , \left [ x , y \right ] \right ] \right ) + { \cal o}(\delta t^4).\end{aligned}\ ] ] here @xmath104 = xy - yx$ ] is the commutator of @xmath101 and @xmath100 . in the case of the baoab method
, we take @xmath105 and @xmath106 to compute the perturbed operator for the method .
a similar analysis can be conducted for the other generalised gla - type methods considered in this paper . to compute @xmath80
we simply plug in our choices into the bch formula to obtain : @xmath107 \right ] + \left[{\cal l}^*_{\rm a } , \left [ { \cal l}^*_{\rm a } , { \cal l}^*_{\rm b } \right ] \right ] \right .
\\ & & \left .
+ \left[{\cal l}^*_{\rm o } , \left [ { \cal l}^*_{\rm a } , { \cal l}^*_{\rm b } \right ] \right ] + \left[{\cal l}^*_{\rm a } , \left [ { \cal l}^*_{\rm o } , { \cal l}^*_{\rm b } \right ] \right ] - \tfrac{1}{2 } \left[{\cal l}^*_{\rm a } , \left [ { \cal l}^*_{\rm a } , { \cal l}^*_{\rm o } \right ] \right ] \right . \\ & & \left .
- \tfrac{1}{2 } \left[{\cal l}^*_{\rm b } , \left [ { \cal l}^*_{\rm b } , { \cal l}^*_{\rm o } \right ] \right ] - \tfrac{1}{2 } \left[{\cal l}^*_{\rm b } , \left [ { \cal l}^*_{\rm b } , { \cal l}^*_{\rm a } \right ] \right ] \right ) + { o}(\delta t^4),\end{aligned}\ ] ] where recall that @xmath108 and @xmath109 the calculation of the inhomogeneity in ( 8) then amounts to a straightforward computation of the commutator series applied to @xmath110 .
the commutators needed are : @xmath111 \right ] \rho_\beta = & 2 \beta p^t u''(x ) \nabla u(x ) \rho_\beta - \rho_\beta \beta \ , p \cdot \nabla_x p^t u''(x ) p , \\
\left[{\cal l}^*_{\rm b } , \left [ { \cal l}^*_{\rm b } , { \cal l}^*_{\rm a } \right ] \right ] \rho_\beta = & -2 \beta p^t
u''(x ) \nabla u(x ) \rho_\beta , \\
\left[{\cal l}^*_{\rm o } , \left [ { \cal l}^*_{\rm a } , { \cal l}^*_{\rm b } \right ] \right ] \rho_\beta = & 2 \gamma \left(\delta_x u(x ) - \beta p^t u''(x ) p \right ) \rho_\beta , \\
\left[{\cal l}^*_{\rm a } , \left [ { \cal l}^*_{\rm o } , { \cal l}^*_{\rm b } \right ] \right ] \rho_\beta = & \gamma \beta \left ( |\nabla u(x ) |^2 - p^t u''(x ) p \right ) \rho_\beta , \\ \left[{\cal l}^*_{\rm a } , \left [ { \cal l}^*_{\rm a } , { \cal l}^*_{\rm o } \right ] \right ] \rho_\beta = & 2 \gamma \left ( \beta |\nabla u(x ) |^2 - \delta_x u(x ) \right ) \rho_\beta , \\
\left[{\cal l}^*_{\rm b } , \left [ { \cal l}^*_{\rm b } , { \cal l}^*_{\rm o } \right ] \right ] \rho_\beta = & 0
\\ \left[{\cal l}^*_{\rm o } , \left [ { \cal l}^*_{\rm o } , { \cal l}^*_{\rm det } \right ] \right ] \rho_\beta = & 0,\end{aligned}\ ] ] where we have abbreviated the hessian @xmath112 hence we see directly that @xmath113 giving @xmath114 . \label{fredalt}\end{aligned}\ ] ] observe that eq .
[ fa ] is satisfied since the average of the first term is equivalent to a canonical average which vanishes @xmath115 whereas the other terms in eq .
[ fredalt ] , being canonical averages of terms which are odd - order in @xmath18 , necessarily also average to zero .
an analogous computation can be performed for the aboba method , giving a slightly ( but crucially ) different perturbation operator @xmath116 , where @xmath117\ ] ] here @xmath118 is the hessian matrix of @xmath41 and @xmath119 is the partial laplacian in @xmath120 .
equation [ fa ] is again seen to be satisfied .
although we are not able to give the general analytical solution to the partial differential equation of eq . [ eqn::pde ] , we can find a solution in an important limiting case : the high friction regime .
to do this , we expand the invariant density of the numerical method further , viewing both @xmath71 and @xmath122 as small parameters : @xmath123 ) .
\label{densityexpansion}\ ] ] dividing by @xmath27 , we may reduce eq .
[ eqn::pde ] to @xmath124(f_{2,0 } + \varepsilon f_{2,1 } + o(\varepsilon^2 ) ) = g_{0 } + \varepsilon g_{1},\ ] ] where @xmath125 and , for baoab , @xmath126 note that this is a singularly perturbed system as @xmath127 is degenerate and it is only the combined operator that has the necessary regularity to define a unique solution . nonetheless , as explained below it is possible to find the leading term @xmath128 by substituting a truncated expansion of fixed degree and solving the resulting equations .
equating powers of @xmath129 , we find @xmath130 and @xmath131 truncating at @xmath132 , for example , we find the following solution of these equations : @xmath133 for aboba , @xmath128 the solution would change to @xmath134 we now turn out attention to the configurational marginal distribution obtained by integrating the density expansion eq .
[ densityexpansion ] with respect to the momenta .
our interest is only in the leading term of this expansion , which defines the sampling behavior for large @xmath27 . ignoring higher order terms in @xmath71 and @xmath129
, we would have the distribution @xmath135 where , for baoab ,
@xmath136 \right ) $ ] , and @xmath137 . noting that the leading term in the exponent of @xmath138 is quadratic in momenta
, we integrate out with respect to @xmath18 to obtain @xmath139 \right ) { \rm d}^n p\\ & = \sqrt{2 \pi k_b t /\det\left ( i + \frac{\delta t^2}{4}u '' \right ) } \exp \left ( -\beta \left [ u - \beta^{-1 } \frac{\delta t^2}{8 } \delta u \right ] \right ) \end{aligned}\ ] ] using the identity @xmath140 , we find @xmath141 \right ) .\ ] ] we then taylor expand the logarithm of @xmath142 and take the trace to obtain a cancellation of the @xmath143 terms , giving @xmath144 the contribution to the configurational distribution error due to @xmath138 is @xmath145 .
this means that the overall error in the marginal distribution of @xmath146 ( which includes the neglected factor @xmath147 ) will be @xmath148 .
if @xmath129 is small ( or @xmath71 is relatively large ) , the error will be dominated by the quartic term in @xmath71 and we will observe 4th order accuracy in configurational averages . for small @xmath71
the method is always eventually second order . in the case of the aboba method
, the remarkable cancellation of the second order errors does not occur and the method always exhibits 2nd order configuration distribution error .
we now consider the limit @xmath149 , where the exact solution of the vector ou process reduces to @xmath150 , where @xmath151 is a vector whose components have a standard normal distribution ( gaussian white noise ) . alternatively
, we could consider the limit of the particle mass going to @xmath152 , although this requires a reformulation of eq .
[ lang ] so that the friction is proportional to the velocity instead of the momentum @xcite .
whichever limit is taken , we would expect the ultimate result to be the same .
( here we have reintroduced the masses in order to present the method , since they may be useful scaling parameters in simulation . ) in the configurations it is straightforward to show that the baoab method therefore becomes @xmath153 where the @xmath13 are vectors of i.i.d .
standard normal random variables . replacing @xmath154 by @xmath16
we arrive at eq .
[ bd - mod ] .
since the langevin scheme gives 4th order accurate configurational averages in this limit , we expect the method of eq .
[ bd - mod ] to be second order accurate in modified timestep @xmath16 .
moreover , since we completely remove the second order term in the langevin dynamics configurational density expansion , we expect to observe this behaviour across all values of @xmath16 .
by contrast the aboba scheme gives a much more complicated limit method as @xmath149 which is not in one - step form . in the euler - maruyama method ,
the random perturbations introduced at each step are independent . in the method of eq .
2 , the random perturbation is a scaling of @xmath155 ; the components @xmath156 of these are independent of each other and decay linearly with timestep : @xmath157 thus , in this new method , we use a colored noise which has characteristics that directly depend on the stepsize , although the noise decorrelates in just a couple of timesteps .
this is therefore no longer a markov process , however it can be reformulated as such if one considers the appropriate extended space ( eg . @xmath158 $ ] ) .
we implemented the methods aboba , baoab , spv and bbk and compared the accuracy of configurational sampling for different values of @xmath27 and a range of timesteps .
a brief analysis shows that the use of the harmonic oscillator leads to special cancellations in the bch series of the splitting schemes , making it a poor test subject .
hence , in order to compare the order of accuracy of the different schemes , we first considered an oscillator model in 1d with potential @xmath159 .
this was accomplished by introducing @xmath11 intervals ( ` bins ' ) of equal length , and computing the mean error in the observed probability frequency compared to the exact expected frequency ( obtained by integration of the probability density ) .
if the observed frequency in bin @xmath160 is @xmath161 , and the exact expected frequency is @xmath162 , then the error calculated is @xmath163 in this one - dimensional example , we used @xmath164 bins to cover the interval from @xmath165 to @xmath166 .
the configurational density error is plotted against stepsize in log - log scale .
if @xmath167 then we expect this graph to be a line of slope @xmath168 .
due to the relative simplicity of this model , we were able to perform highly resolved simulations to calculate accurate error estimates for the configurational distribution .
the exact expected value that we compare experimental results against can be computed to arbitrary prescision , and we are able to run as many simulations as needed in order to drive the variance of results to a minimum . the variance in our results , in cases where the stepsize was less than @xmath169 , was consistently below @xmath170 . above a stepsize of @xmath169 some of the methods
were found to be unstable .
the results of our simulations are summarized in figure [ quartic ] . .
the simulation time was fixed for all runs at @xmath171 , and five runs were averaged to further reduce sampling errors . at left , @xmath172 , at right @xmath173 .
the graphs are entirely in keeping with the theory presented in the article .
[ quartic],title="fig:",width=288 ] .
the simulation time was fixed for all runs at @xmath171 , and five runs were averaged to further reduce sampling errors . at left , @xmath172 , at right @xmath173 .
the graphs are entirely in keeping with the theory presented in the article .
[ quartic],title="fig:",width=288 ] as we can see , when @xmath27 is small , the methods perform somewhat similarly , at least in the qualitative sense , with all showing a 2nd order error in configurational sampling , and the aboba and spv methods essentially identical . as @xmath27
is increased , the substantial difference between baoab and the other methods becomes apparent . in the limit of large
@xmath27 the spv method effectively annihilates the force which results in poor sampling . in the graph for @xmath172
, we can see that for larger values of the stepsize , the graph steepens ( indicating that the fourth order term is dominant ) ; as the stepsize is decreased the method exhibits a 2nd order asymptotic decay .
with @xmath173 the fourth order behavior is seen for for all indicated data points , although , again , this becomes second order for smaller values of @xmath71 .
note that the limit method eq . 2 ( with the substitution @xmath174 )
gives an essentially identical behavior to the @xmath173 case .
we also give a comparison of the actual computed configurational distributions , at different stepsizes using each method , for @xmath175 in figure [ 1ddist ] . .
the three curves for each of the four methods show the results for three different stepsizes : @xmath176 ( circles ) , @xmath177 ( crosses ) , @xmath178 ( dashed ) , compared to the dark , solid curve representing the exact distribution .
the graph shows that the generalized gla methods are superior to spv and bbk in the moderate @xmath27 regime .
[ 1ddist],width=576 ] , for morse ( left ) and lennard - jones ( right ) clusters .
they also show the choice of bins used in calculating the numerical distributions .
[ pic::gofr],title="fig:",width=288 ] , for morse ( left ) and lennard - jones ( right ) clusters .
they also show the choice of bins used in calculating the numerical distributions .
[ pic::gofr],title="fig:",width=288 ] to examine the performance of the limit method in more detail , we next considered small molecular clusters consisting of seven atoms ( motion restricted to the plane ) , with both morse ( @xmath179 ) and lennard - jones ( @xmath180 ) potentials , given by @xmath181 where we use @xmath182 and @xmath183 .
the overall potentials are hence @xmath184 where @xmath185 is the distance between particles @xmath160 and @xmath186 , @xmath187 is the distance between particle @xmath188 and the origin , and a mild harmonic term is included in the case of the lennard - jones system in order to prevent particles being ejected from the cluster .
the morse potential gives a smoother dynamics compared to lennard - jones ( the morse forces were on average three times smaller than those for lennard - jones ) and allows a more satisfying determination of the error scaling behavior with stepsize . to quantify the error in configurational sampling , we calculated the radial density @xmath189 by binning the instantaneous interatomic distances at each step into 20 compartments and compared to a calculated reference value ( figure [ pic::gofr ] ) .
though not exact , the errors in the reference will be negligible compared with configurational distribution errors at higher stepsizes . for the morse cluster the reference stepsize we used was @xmath190 , whereas for the lennard - jones cluster , we used @xmath191 both with the same integration time that was used in their respective test runs . in both cases
the cluster was initialized with 6 particles placed on a regular hexagon with unit side - length and with the remaining particle in the centre , and initial velocities randomly drawn from the canonical distribution .
considerable computation is required to achieve the level of accuracy required , due to the dominance of sampling error and the complexity of the system compared to the earlier 1d example .
we ran a large number of independent simulations at each stepsize and computed the average radial density plot for both examples , and compared this to the reference result .
our results , presented in figure [ morse ] , are entirely consistent with our analysis and show the second order dependence of the configurational sampling error on @xmath16 ( equivalent to fourth order in @xmath71 ) , as compared to the euler - maruyama s first order behavior .
these results demonstrate a good agreement with our theoretical results , however even using extensive computation the variances in each experiment were still quite high .
if @xmath192 is the density of bin @xmath193 in simulation @xmath194 at stepsize @xmath16 , we calculate the variance as @xmath195 where we used @xmath196 bins for each experiment , @xmath197 for the morse experiment and @xmath198 for the lennard - jones experiment .
the variances for the morse experiment were around @xmath170 while the lennard jones experiment had variances around @xmath199 .
we expect that completing more simulations ( increasing @xmath0 ) would reduce the variance and give smoother results in figure [ morse ] .
nonetheless , in both cases , we observe a significant reduction in the error compared to euler - maruyama . ) . left : morse potential ; right : lennard - jones . for morse we used a temperature of @xmath200 , a fixed time interval of @xmath201 , with stepsizes ranging from @xmath202 to @xmath203 . for lennard - jones
the temperature was @xmath204 , @xmath205 and stepsizes ranged from @xmath206 to @xmath207 . in order to drive the variance of the results down , a large number of runs were necessary : for the morse simulation the error is computed using the average histogram computed from 200 independent runs , while the lennard - jones simulation used 1000 independent runs . around two orders of magnitude of improvement are observed in the accurate regime , but , perhaps even more important , the baoab limit method is usable at substantially larger stepsizes than euler - maruyama .
[ morse],title="fig:",width=288 ] ) . left : morse potential ; right : lennard - jones .
for morse we used a temperature of @xmath200 , a fixed time interval of @xmath201 , with stepsizes ranging from @xmath202 to @xmath203 . for lennard - jones
the temperature was @xmath204 , @xmath205 and stepsizes ranged from @xmath206 to @xmath207 . in order to drive the variance of the results down , a large number of runs were necessary : for the morse simulation the error is computed using the average histogram computed from 200 independent runs , while the lennard - jones simulation used 1000 independent runs . around two orders of magnitude of improvement are observed in the accurate regime , but , perhaps even more important , the baoab limit method is usable at substantially larger stepsizes than euler - maruyama .
[ morse],title="fig:",width=288 ] it might be a suggested that the improved accuracy seen in the high @xmath27 regime could potentially come at the price of a slower convergence to equilibrium due to a reduced rate of transition between metastable states , hence overall sampling of the configurational distribution might be impaired in favour of local sampling . to address this point ,
we have plotted in figure [ rateofconv ] the error computed in our lennard - jones simulation as a function of the number of force evaluations ( vertical ) against the friction value @xmath27 used for the simulation ( horizontal ) .
gridpoints in the plot are coloured according to the configurational error , computed using eq .
[ errorformula ] .
the results indicate that the convergence rate for the baoab method is not diminished for large friction coefficient , so it does not appear that we are sacrificing sampling accuracy using this scheme .
note that the performance of aboba is also robust in the limit of large @xmath27 , but the achievable accuracy is reduced as it is only second order , consistent with what we have presented in this article .
our results confirm the theoretical results of sections 3 and 4 , in particular showing the higher order configurational sampling of the baoab method .
the order in its langevin formulation is effectively four in the large @xmath27 limit and large values of @xmath27 do not impair its stability due to the use of exact ornstein - uhlenbeck solves . not only is the order of the method high , the constant multiplying the leading term must be , in the cases looked at here , of modest magnitude , since the errors are relatively small also at the large stepsize stability threshold .
since , in the context of molecular dynamics , baoab is a ` cheap ' and easy to implement scheme using only a single force vector per timestep , we stress that there is no price to pay for its improved accuracy . in molecular modelling
, there are other errors that play important roles , most importantly errors in the force fields ( or , more fundamentally , the errors due to not modelling quantum mechanics properly ) and sampling errors . obviously these errors may dominate the overall method error and limit the relative benefit to be gained by using one integrator as compared to another , but it is also clear that both of the other types of errors are constantly being reduced through the design of better models and the use of more powerful computers .
more important , one can ask the question : how can a practitioner know which part of the error in a given complicated simulation is due to sampling error and which part due to the truncation errors addressed here ? in our experiments with molecular models , even where there was substantial sampling error still present , we nonetheless found the accuracy to be noticeably higher for the baoab method ; it is likely that this improvement in sampling accuracy would be of direct benefit in many real world simulations .
finally we point out that the baoab scheme and its limit method ( eq .
2 ) was , in each case studied , stable at a larger stepsize than the alternatives , meaning that longer time intervals are made accessible .
this was particularly dramatic in the case of the lennard - jones system .
* acknowledgements . *
the first author acknowledges the support of a jto fellowship from the institute for computational engineering and sciences at the university of texas .
the second author was supported by the centre for numerical algorithms and intelligent software ( funded by epsrc grant ep / g036136/1 and the scottish funding council ) .
a. stuart , h. owhadi , and especially g. stoltz made helpful suggestions regarding the presentation of the results .
we further acknowledge the comments received from a. abdulle and m. tretyakov which have improved the article and its relation to other works .
computations were performed on the university of edinburgh s ecdf compute facility .
99 brnger , a , brooks iii , c , & karplus , m. ( 1982 ) stochastic boundary- conditions for molecular - dynamics simulations of st2 water .
letters _ * 105 * : 495500 .
talay , d.(1995 ) simulation and numerical analysis of stochastic differential systems : a review . in _
probabilistic methods in applied physics _ , p. kree and w. wedig ( eds . ) , vol .
451 of lecture notes in physics , pp .
springer - verlag .
mishra , b. and schlick , s.(1996 ) the notion of error in langevin dynamics .
i. linear analysis _ j. chem .
* 105*:299318 .
milstein , g. and tretyakov , m. ( 2004 ) _ stochastic numerics for mathematical physics _
, springer .
burrage , k. and lythe , g.(2009 ) accurate stationary densities with partitioned numerical methods for stochastic differential equations .
_ siam j. numer .
* 47*:16011618 .
lelivre , t. , rousset , m. , & stoltz , g. ( 2010 ) _ free energy computations . a mathematical perspective .
_ imperial college press .
liu , j. ( 2001 ) _ monte carlo strategies in scientific computing .
_ springer series in statistics .
larini , l. mannella r. , & leporini , d. ( 2007 ) langevin stabilization of molecular - dynamics simulations of polymers by means of quasisymplectic algorithms .
_ j. chem .
phys . _ * 126*:104101 .
skeel , r and izaguirre , j. ( 2002 ) an impulse integrator for langevin dynamics _ mol .
phys . _ * 100*:38853891
. milstein , g. and tretyakov , m. ( 2003 ) quasi - symplectic methods for langevin - type equations _
i m a j. num .
_ * 23 * : 593626 .
shardlow , t. , ( 2006 ) modified equations for stochastic differential equations .
_ bit _ * 46 * , 111125 melchionna , s.(2007 ) design of quasi - symplectic propagators for langevin dynamics .
_ j. chem .
_ * 127*:044108 .
bou - rabee , n. and owhadi , h. ( 2010 ) long - run accuracy of variational integrators in the stochastic context .
_ siam j. num .
* 48 * : 278297 .
davidchack , r.l.(2010 ) discretization errors in molecular dynamics simulations with deterministic and stochastic thermostats .
_ j. comput .
phys . _ * 229 * : 9323 9346 .
zygalakis , k. ( 2011 ) on the existence and applications of modified equations for stochastic differential equations .
_ siam j. sci .
* 33*:102130 .
hairer , e. lubich , c. & wanner , g. ( 2006 ) _ geometric numerical integration_. springer ( new york ) , second edition .
leimkuhler , b. and reich , s. ( 2005 ) _ simulating hamiltonian dynamics _ , cambridge university press .
frenkel , d. and smit , b. ( 2001 ) _ understanding molecular simulation _
, 2nd edition , academic press .
hrmander , l. ( 1967 ) hypoelliptic second order differential equations . _ acta math . _ * 119 * : 147171 .
mattingly , j.c .
stuart , a.m. & higham , d.j.(2002 ) ergodicity for sdes and approximations : locally lipschitz vector fields and degenerate noise .
appls . _ * 101*:185232 .
villani , c. ( 2009 ) hypocoercivity .
soc . _ * 202 * pavliotis , g. and hairer , m. ( 2008 ) from ballistic to diffusive behavior in periodic potentials .
_ j. stat .
phys . _ * 131 * 175202 joubaud , r. and stoltz , g. ( 2012 ) nonequilibrium shear viscosity computations with langevin dynamics .
_ siam mms _ * 10 * : 191216 sivak , d. chodera , j. and crooks , g. talay , d. ( 2002 ) stochastic hamiltonian dissipative systems : exponential convergence to the invariant measure , and discretization by the implicit euler scheme .
_ markov processes and related fields _ * 8*(2 ) 163198 .
mattingly , j.c .
stuart , a.m. & tretyakov , m. ( 2010 ) convergence of numerical time - averaging and stationary measures via poisson equations .
_ siam j. num .
* 48 * : 552577 .
debussche , a. and faou , e.
the langevin dynamics methods used for the numerical experiments in this paper are given here .
@xmath208 is an @xmath209vector of i.i.d .
normal random numbers , with mean @xmath152 and variance @xmath210 .
the diagonal mass matrix is denoted @xmath11 , and we assume a timestep @xmath71 is provided . given the parameter @xmath27 , we define useful constants @xmath211 and @xmath212 | in this article , we focus on the sampling of the configurational gibbs - boltzmann distribution , that is , the calculation of averages of functions of the position coordinates of a molecular @xmath0-body system modelled at constant temperature .
we show how a formal series expansion of the invariant measure of a langevin dynamics numerical method can be obtained in a straightforward way using the baker - campbell - hausdorff lemma .
we then compare langevin dynamics integrators in terms of their invariant distributions and demonstrate a superconvergence property ( 4th order accuracy where only 2nd order would be expected ) of one method in the high friction limit ; this method , moreover , can be reduced to a simple modification of the euler - maruyama method for brownian dynamics involving a non - markovian ( coloured noise ) random process . in the brownian dynamics case
, 2nd order accuracy of the invariant density is achieved .
all methods considered are efficient for molecular applications ( requiring one force evaluation per timestep ) and of a simple form .
in fully resolved ( long run ) molecular dynamics simulations , for our favoured method , we observe up to two orders of magnitude improvement in configurational sampling accuracy for given stepsize with no evident reduction in the size of the largest usable timestep compared to common alternative methods .
* keywords : * molecular dynamics ; sampling ; langevin dynamics ; brownian dynamics ; stochastic dynamics . |
quantum computation is of great interest in recent years for its great speed up for solving some problems@xciteand its efficiency in simulating physical systems@xcite .
ideally , a quantum computer is a closed quantum system , and the evolution of the system is a unitary operation .
however , as any quantum computer is inevitably interacted with the environment , the system is more or less open , and thus the system evolution becomes non - unitary .
this is one of major difficulties for making a large scale quantum computer .
when a quantum system becomes open , the information about the initial state may be lost after the evolution of the system , and we can not undo a quantum operation in general .
this is discussed in quantum commutation , where quantum channel capacity is introduced to describe the information transfer ability of a quantum channel ( see , e.g. , ref .
quantum channel capacity also illustrates the quantum channel noise .
notably , there are also other ways to study the noise of a quantum channel , such as using the addition of noise to change a quantum map into an entanglement - breaking map @xcite .
since we can divide quantum operations into two categories : the unitary operations which are ideal , and the non - unitary ones which are most of the cases in real quantum systems , it is nature to ask to what extend a general quantum operation deviates from the group of unitary quantum operations , which appears to be fundamentally important .
we here treat all of the unitary operations as a group because the information of a general quantum state can be perfectly preserved for all unitary operations . to study the distance between a given quantum operation and the group of unitary operations , a common way is to find the closest unitary quantum operation for the given one , and use the distance between these two operations as the distance between the given quantum operation and the unitary operation group .
the measure of the distance between two quantum processes has already been well studied@xcite , and we choose quantum process matrix fidelity in this paper .
we define the fidelity between a given quantum process and its closest unitary one as the degree of unitarity(du ) of this given quantum process .
we think the study of the measure of the du is important because of the following reasons .
first of all , du quantifies the difference between a realistic quantum operation and the ideal ones .
it is an important intrinsic property for a quantum operation .
also , the du of a quantum process is closely related to the noise of this process , the quantum capacity , and some other physical quantities such as the non - markovian behavior of a quantum process . to obtain the du of a given quantum process ,
a core problem we need to solve is to find the closest unitary operations of the given quantum process , since an analytical result for the measure of the distance between two quantum processes has already been available @xcite . however , finding an optimal unitary matrix is in general difficult because the optimization should be taken over all the unitary matrices .
remarkably , we here obtain analytical results for all qubit unital channels . moreover , we also reveal the upper and lower bounds for the du of a general quantum process .
the lower bound is very tight and deviates from the true value slightly for most quantum processes , and it can be treated as a good approximation of the du in most cases . our paper is organized as follows . in the next section ( sec .
ii ) , we give the definition of the du of a quantum process and address its properties . in sec .
iii , we give analytical results for qubit unital channels and some other special cases . in sec .
iv , we reveal the upper and lower bounds for the du of a quantum process in general cases , and discuss the tightness of the bounds . in sec .
v , we study the probability distribution of du of a quantum system interacting with environment of different dimensions .
finally , we present a brief summary and outlook in sec . vi .
a quantum process @xmath0 can be expressed in kraus operators@xcite , @xmath1 where the operators satisfy @xmath2 to insure this is a trace preserving map .
the kraus operator representation has clear physical meaning and every part of the summation can be treated as the evolution of the system when the environment is measured and a specific result is obtained .
but since we need a group of matrices to represent a quantum process using kraus operators , sometimes it is more convenient to use quantum process matrix to describe a quantum process which only need one matrix .
suppose @xmath3 is an orthogonal basis set for the system , let @xmath4 , then the quantum process @xmath0 can be expressed as @xmath5 where @xmath6 is called the quantum process matrix which contains all the information of the quantum process . for a given quantum process @xmath0 and a unitary operation @xmath7 , how to measure
the distance between them has been widely discussed@xcite .
one way is consider the difference of the out put states for @xmath7 and @xmath0 . as we want the distance between two processes to be independent of the initial state ,
a nature way is to average over all the initial states .
if we choose fidelity to measure the difference of the output states , the distance between the processes can be measured by @xmath8 here @xmath9 is the initial state of the system . the integration is carried over the whole hilbert space of the quantum state of the system , and the volume of the state is calculated according to harr measure .
the average fidelity measure not only has clear physical meaning , but also has analytical expression . in ref.@xcite
, they give an analytical result in this formula : @xmath10 where @xmath11 is the dimension of the system . another way to measure
the distance is to measure the distance between the quantum process matrices directly .
we also choose fidelity to measure the distance between process matrices .
then the fidelity between two processes is @xmath12 where @xmath13 .
it turns out that @xmath14the average fidelity measure of two quantum processes and the fidelity between two process matrices is related by the formula below@xcite : @xmath15 substituting it into eq .
( [ eq : average fidelity ] ) , we can get that @xmath16 we choose @xmath17 to measure the distance between the quantum process @xmath0 and the unitary evolution @xmath7 in this paper .
the reasons for this will be stated later in this section . in the following ,
we denote @xmath17 as @xmath18 for simplicity .
now we consider the main topic of this paper , the measure of the degree of unitarity(du ) for a given quantum process @xmath0 . as it is discussed in sec .
[ sec : introduction ] , we choose the nearest unitary operation @xmath19 for the given quantum process @xmath0 and use @xmath20 to measure the distance between @xmath0 and @xmath7 .
we define @xmath20 as the du for @xmath0 .
this definition can be treated as the geometry measure of the du , and the geometry measure is also used in other cases such as quantum discord@xcite . according to the definition and eq .
( [ eq : processfidelity ] ) , the du of a quantum process @xmath0 is , @xmath21 generally , it is hard to get analytical result for the above expression as the optimization is over all the unitary matrices .
luckily in this problem for some special cases we can get clean result . we will discuss this in details in the next section .
here we focus on the properties of du to illustrate that this is indeed a reasonable definition and measures the similarity between a general quantum process and unitary ones .
first , the @xmath22 of @xmath0 should be a property of the process and should not be depended on the choice of kraus operators @xmath23 .
this property is assured by that the average fidelity is independent of the choice of the kraus operators@xcite .
also , the du of @xmath0 should not be smaller than the du @xmath24 , which means that the quantum process can not become more unitary by subsequent processes .
this is also guaranteed by the properties of average fidelity@xcite .
this property also make the du as an indicator for non - markovian behavior of a quantum process@xcite .
it can be deduced that if du increases during a quantum evolution , then this quantum evolution is non - markovian . however , the converse is not true .
next we consider the extreme values of the du of @xmath25 .
when @xmath0 is unitary , the du should reach its maximum .
this can be easily confirmed because in this case the nearest unitary operation is @xmath0 itself and one can get that du of this unitary process is 1 . for the minimum of du , intuitively
, one can choose the quantum process(or quantum channel ) to be a maximal depolarizing channel , which maps all the initial states into the maximum mixed state .
in this case , one can choose arbitrary unitary operation as the nearest one , and the du of this process is equal to @xmath26 .
finally we discuss the difference between the du of @xmath0 and @xmath27 , where @xmath7 is a unitary operation acting on an additional quantum system .
one can verify that @xmath28 which means that the du of a given quantum process will not change by adding an unitary operation imposed on an ancillary quantum systems . also one can verify that if we choose the average fidelity as the measure for two quantum processes , although the nearest unitary operation will not change , the above property will not be satisfied .
this is the reason that we choose the process matrix fidelity measure .
one can also refer to table 1 to get a clearer picture of du , where the du of some important qubit channel is listed .
table 1 can be get using the method provided in the next section .
to get the du of a quantum process
@xmath0 , we need to find the nearest @xmath7 from the group of unitary matrices .
this is in general difficult .
however , if the quantum process @xmath0 satisfies some properties , we can get analytical result .
we introduce the inner product in the operator space , and denote the inner product between two operators @xmath29 and @xmath30 as @xmath31 : @xmath32 then according to eq .
( [ eq : definition of du ] ) , @xmath14the du of the quantum process @xmath0 can be expressed as @xmath33 so the du of a quantum process is the sum of the square of the projections of the nearest @xmath7 on each kraus operator .
if each kraus operator is a unitary operation multiplied by a constant , and they are orthogonal with each other , then the optimization process can be greatly simplified .
_ theorem 1 : _ for a quantum process @xmath34 , if @xmath35 for all @xmath36s and @xmath37 for @xmath38 , then @xmath39 where @xmath40 is the coeficient with the max norm .
_ proof : _
expand @xmath41 into a set of orthogonal basis in the operator space with operations @xmath42 .
then any unitary operation @xmath7 can be expressed as a linear combination of @xmath43 , @xmath44 where @xmath45 .
then @xmath46 suppose @xmath47 is the coefficient with the biggest norm .
as @xmath48 , one can see that the above expression reaches its maximum when @xmath49 . in this case
, according to the expression of the du of a quantum process in eq .
( [ eq : definition of du])@xmath14 , @xmath50 this completes the proof . of course , the situation in theorem 1 is a very special case , it requires all all the kraus operators to be proportional to unitary operations and are orthogonal to each other at the same time .
but one can notice that the representation of a quantum process using kraus operators has its freedom and it turns out that every quantum process can be represented in orthogonal kraus operators@xcite .
suppose a quantum process @xmath34 , let @xmath51 be the correlation matrix with @xmath52 .
one can note that @xmath51 is a hermitian matrix .
we can diagonolize @xmath51 with unitary matrix @xmath53 @xmath54 it can be proved that the rank of @xmath55 is @xmath56 at most , where @xmath11 is the dimension of the system .
let @xmath57 one can show that @xmath58 is equivalent with @xmath59 and @xmath60 further more , for a qubit channel , if the original kraus operators @xmath23s are proportional to unitary operations , then after diagonaliztion , the operators @xmath61 are still unitary operations .
this theorem can be found in ref.@xcite .
here we give a different proof using the notations in this paper .
_ proof : _
suppose the kraus operators are @xmath62 , where @xmath63 , @xmath64 .
follow the diagonalization process stated above , we can get the kraus operators after diagonalization @xmath65 .
then @xmath66 one can note that if @xmath23 is replaced by @xmath67 with @xmath68 , the quantum process will be the same as before .
so we can treat all the @xmath69s as su(2 ) matrices . for a su(2 )
matrix , one can represent it as @xmath70 one can verify that @xmath71 is a real , and @xmath72 also one can notice that @xmath51 is an orthogonal matrix in this case , so @xmath53 is real too . on the r.h.s . of eq .
( [ eq : fkdaggerfk ] ) , if @xmath73 , @xmath74 is proportional to @xmath75 . if @xmath76 , as @xmath53 is real and @xmath77 is proportional to @xmath75 , @xmath78 is also proportional to @xmath75 .
so we can get @xmath79 where @xmath80 is a complex number .
this condition implies that @xmath61 is proportional to a unitary operation .
combined with theorem 1 together , we can get that if a qubit channel is a convex combination of unitary channels , we can get analytical result for the du of this channel . for qubit channels
, it can be shown that every unital channel is a convex combination of unitary channels , where a unital channel is a channel that maps @xmath81 into @xmath82 .
this is related to birkhoff conjecture , which is only true for @xmath83@xcite .
since we can get analytical result for convex combination of unitary channels , we can get analytical result of the du for all qubit unital channels .
convex combination of unitary channels are of great interest as they are the only channels that can be perfectly inverted by monitoring the environment@xcite .
the calculation of du for these channels offers a new perspective for the information and the noise of these channels .
now we list the analytical result of du in table 1 for some important qubit channels , including depolarizing channel , bit flip channel , phase flip channel , and amplitude damping channel .
we can see that the du of these quantum channel is between @xmath84 and 1 , as it is shown in sec .
[ sec : definition - and - properties ] . for depolarizing channel , bit flip channel and phase flip channel , the nearest unitary operation changes when the parameter of the channel changes .
for amplitude damping channel , the nearest unitary operation is always the identity operation . [
cols="^,^,^",options="header " , ]
in general cases , it is hard to get analytical result for the du of a quantum process .
we try to get the lower bound and upper bound for du in this section .
to get the lower bound , we try to find a unitary operation that is very close to the quantum process , although it may not be the closest . as it is hard to find the closest @xmath7 for all these kraus operators , we try to find the closest @xmath7 for one of them .
one option is to find the nearest @xmath7 for the kraus operator with the largest norm after diagonalization .
suppose the kraus operators of @xmath0 after diagonalization are @xmath58 , and the norm of @xmath85 , @xmath86 , is the largest among all the @xmath87s .
then the closest unitary matrix for @xmath85 , which is the unitary matrix that maximize @xmath88 , can be found though polar decomposition .
suppose the polar decomposition of @xmath85 is @xmath89 then @xmath90 where the equality is obtained when @xmath91 .
one can note that @xmath92 is the nearest unitary matrix for @xmath85 .
we choose @xmath92 as the approximation of the nearest unitary operation for @xmath0 , and calculate the du of this quantum process .
then we get the lower bound for du of @xmath0 , @xmath93 also as @xmath87s are orthogonal with each other , the contribution of @xmath94 is relatively small if @xmath95 is not 1 , i.e. , if @xmath87 is not the kraus operator with the largest norm .
so we can also choose @xmath96 as a simplified lower bound , which equals @xmath97 , where @xmath98s are the singular value of @xmath85 .
this lower bound equals the largest visibility the interference between @xmath0 and unitary channels , and it can be got experimentally@xcite .
we can get the lower bound in another way .
we choose the kraus operator that may contribute most to @xmath99 as the major part , of which the value @xmath100 is the maximum .
one can note that chosen kraus operator has the largest value @xmath14of @xmath101 . @xmath14note
that previously we choose the one with the largest value of @xmath102@xmath14 .
this two standards may make different when the derivation of @xmath98s is large .
we denote the major kraus operator by this new standard as @xmath103 , and the nearest unitary operation of @xmath103 as @xmath19 , then the new low bound for @xmath99 is @xmath104 now we consider the upper bound of the du of a quantum process . we assume the nearest @xmath7 of @xmath0 is the nearest unitary operation of all the kraus operator @xmath87s , which means that there is a unitary operation that makes @xmath100 reaches its largest value for all @xmath87s .
apparently , this is not possible for most cases .
but it give a upper bound for the du of this quantum process .
according to eq .
( [ eq : maxinnerproduct ] ) , the largest value of @xmath100 is @xmath14@xmath101 .
so we can get a upper bound for @xmath99 , @xmath105 we can show that the upper bound of @xmath99 is less or equal than 1 . as @xmath106
now we consider the tightness of the lower bound and upper bound .
we denote the lower bound in eq .
( [ eq : lowerbound1 ] ) as lower bound 1 , and the one in eq .
( [ eq : lowerboud2 ] ) as lower bound 2 .
we randomly choose a quantum process and calculate the du of the quantum process , two lower bounds of the du , and the upper bound of du . also , in order to study the bound behavior as a function of the du , we choose approximately equal amount of quantum processes for each interval of the du of quantum processes .
as shown in fig .
1(a ) , the lower bound of du(red stars ) is a good approximation of du(black cross ) in almost all the cases .
particularly , when du is larger than 0.4 , the lower bound of du(red stars)is almost the same as du .
when the du of a quantum process is very small , the lower bound may differs from the true value for about 10% .
on the other hand , the upper bound of du is quite loose .
but it can be shown that the upper bound may be a indicator of the behavior of the lower bound .
as shown in fig .
1(b ) , the difference of two lower bounds of du from the true value are almost zero when the upper bound of du is bigger than 0.8
. when the upper bound of is small , the error of the lower bound may be large .
in this section , we consider the probability distribution of the du of a random quantum process for a quantum system .
this distribution can give us some information for the du of a random quantum process before we really know it .
it also gives the expected value of du of the process of a quantum system .
one can note that the distribution of du depends on the dimension of the system and the environment .
it is also related to how the system and the environment interacts with each other . here
we study a qubit system .
suppose the qubit is fully interacting with an environment with an dimension of @xmath107 , and the evolution of the total system composed by the qubit and the environment is a random unitary operation with the dimension of @xmath108 .
in this case , we study the probability distribution of du of the operations imposed on the qubit system . for simplicity
, we only consider the cases when @xmath109 .
we generate a random unitary operation for the total system according to the harr measure , and calculate the du for the corresponding quantum operations imposed on the principle qubit system . as it is already tested in sec .
[ sec : lower - boudn - and ] that the lower bound of du is very tight , we treated the lower bound of du as du in this calculation .
as shown in fig .
2 , when the dimension of the system become bigger , the expected value for the du is decreased .
this means that the system is expected to be more open and the evolution of the system is expect to be more non - unitary .
in conclusion , we have investigated a key problem of the du for any quantum process .
we have introduced a definition of the du of a quantum process and addressed its properties .
the du of a quantum process quantifies the distance between a given quantum process and the group of all unitary ones .
it is closely related to the noise of the quantum process and is an indicator for non - markovian behavior of a quantum system . for qubit unital channels
, we have obtained an analytical result of the du . for general cases ,
we have derived the lower and upper bounds for the du .
we have presented two different lower bounds for the du , both being quite tight in most cases .
the upper bound of du can be treated as an indicator for the tightness of du .
when the upper bound is low , the lower bound of du may be less tight .
we have also discussed the probability distribution of the du of a qubit system interacting with different environments , and found that the du tends to become smaller when the dimension of the environment become bigger .
the study of du of a quantum process raises many related issues to be investigated in near future , such as the relationship between du of a quantum process and the noise of a quantum process according to other measures , the du s evolution behavior for a real physical system , the du for the quantum process in an open quantum system based quantum algorithm@xcite and so on .
karl kraus , a. bo hm , j. d. dollard , and w. h. wootters , states , effects , and operations fundamental notions of quantum theory : lectures in mathematical physics at the university of texas at austin , lecture notes in physics , vol .
190 ( springer , berlin , 1983 ) . | quantum processes can be divided into two categories : unitary and non - unitary ones . for a given quantum process
, we can define a _ degree of the unitarity ( du ) _ of this process to be the fidelity between it and its closest unitary one .
the du , as an intrinsic property of a given quantum process , is able to quantify the distance between the process and the group of unitary ones , and is closely related to the noise of this quantum process .
we derive analytical results of du for qubit unital channels , and obtain the lower and upper bounds in general .
the lower bound is tight for most of quantum processes , and is particularly tight when the corresponding du is sufficiently large .
the upper bound is found to be an indicator for the tightness of the lower bound .
moreover , we study the distribution of du in random quantum processes with different environments . in particular , the relationship between the du of any quantum process and the non - markovian behavior of it is also addressed . |
the star formation history of field stars in the large magellanic cloud ( lmc ) contains much information about the formation and dynamics of our closest galactic neighbor .
studies by stryker ( 1984 ) , bertelli _ et al . _ ( 1992 ) , westerlund _
et al . _ ( 1995 ) , and vallenari _ et al . _
( 1996a , b ) , among others , conclude that the lmc field contains a majority of young to intermediate age stars overlying a minority old population .
et al . _ ( 1992 ) favor a star formation history in which the star formation rate , initially at a constant low level , increases by a factor of ten in the last few billion years .
this star formation history produces a stellar population reminiscent of the bimodal age distribution of lmc globular clusters ( van den bergh 1991 ; girardi _ et al . _
_ ( 1996b ) find tentative evidence that the onset of this increase in star formation is correlated with position in the lmc , and suggest that this correlation might arise if star formation is triggered by tidal interactions with the small magellanic cloud .
et al . _ ( 1997 ) present _ hst _ observations of a field in the bar of the lmc and find evidence for an additional younger population of stars which is not observed in the outer regions of the lmc field .
a clearer understanding of stellar populations throughout the lmc should provide clues about the age and formation history of the lmc as well as about the mechanisms which trigger star formation in this galaxy .
we have observed three fields in the lmc with the wide field planetary camera 2 on the _ hubble space telescope _ to determine whether these regions share a similar formation history .
the fields are all located in the outer regions of the lmc at roughly the same radial distance from the lmc bar .
these observations extend several magnitudes below the main sequence turnoff and provide a significant improvement over ground - based studies .
gallagher _ et al . _
( 1996 ) present the color - magnitude diagram for one of these fields , and find that the width of the upper main sequence is consistent with a star formation rate which is roughly constant for the last few billion years .
in addition , they suggest that a small burst of star formation occurred 2 gyr ago , leading to a distinct subgiant branch seen one magnitude brighter than the faintest main sequence turnoff .
the lack of evidence in the _ hst _ data for a strong star formation burst is in apparent contrast to previously determined star formation histories and results which indicate the metallicity of the lmc has nearly doubled in the past 2 gyr ( dopita _ et al . _
1997 ) .
holtzman _ et al .
_ ( 1997 ) analyze the luminosity function of the same _ hst _ field to constrain its initial mass function and star formation history . by comparing the luminosity function of the lower main sequence to stellar models ,
they constrain the imf slope , @xmath0 ( @xmath1 ) , to @xmath2 in the mass range @xmath3 . assuming a salpeter imf ( @xmath4 )
, they derive a star formation history from the entire observed luminosity function .
they favor a star formation history in which the star formation rate is roughly constant for 10 gyr and then increases by a factor of three for the past 2 gyr , resulting in a stellar population with comparable numbers of stars older and younger than 4 gyr .
this is in contrast to bertelli _
et al . _ ( 1992 ) , whose preferred star formation history produces a predominantly young ( @xmath5 gyr ) stellar population .
_ find that a predominantly young population fits the _ hst _ observations only if the imf slope is steeper , with @xmath6 . in @xmath7
, we present _ hst _ observations of the three lmc fields . in @xmath8 , we quantitatively compare the stellar populations in these fields and show that they are statistically indistinguishable . in @xmath9 , we compare our observations with several possible star formation histories , using the r - method of bertelli _ et al . _ ( 1992 ) in combination with comparisons between the model and observed luminosity functions .
our derivation of the star formation history is an improvement over previous _
determined formation histories as we use ground - based data to supplement observations at the brightest magnitudes .
observations were made with the wide field planetary camera 2 of the _ hubble space telescope _ between may 1994 and december 1995 through the f555w ( @xmath10 ) and f814w ( @xmath11 ) filters .
total exposure times were 4000s , 2500s , and 1000s in each filter for fields 1 , 2 , and 3 , respectively .
observations through each filter were split into three or more separate exposures to allow identification and removal of cosmic ray events . in figure 1 ,
an image of the lmc shows the approximate positions of the three fields .
a previous analysis of field 1 has been presented by gallagher _
et al . _ ( 1996 ) and holtzman _ et al . _
( 1997 ) .
the positions and exposure times for each field are listed in table 1 .
the data were processed using standard reduction techniques described in holtzman _
et al . _ ( 1995a ) .
this process includes a small correction for analog - to - digital errors , overscan and bias subtraction , dark subtraction , a small shutter shading correction , and flat fielding . in each filter , the images were combined and cosmic ray events were removed based on the expected variance from photon statistics and read noise .
a combination of profile - fitting and aperture photometry was chosen to give good photometry at both bright and faint signal levels .
since the f555w and f814w images were roughly equally deep , stars were found for each field on the summed frame of these two images .
due to structure in the point spread function ( psf ) , objects found in the area surrounding the peak of bright stars were rejected . using this star list ,
profile - fitting photometry was performed on each frame .
the model psfs were the same as those used by holtzman _
et al . _ ( 1997 ) .
profile - fitting results were then used to subtract all stars from the images .
final magnitudes were determined by adding each star individually back into the subtracted frame and performing aperture photometry with a 2 pixel radius aperture .
aperture corrections to a 0.5 arcsecond radius aperture were individually determined for the four wfpc2 chips from bright , isolated stars .
we estimate the maximum error of this correction to be a few hundredths of a magnitude .
instrumental magnitudes were transformed into v and i magnitudes using the transformations given by holtzman _
et al . _ ( 1995b ) .
to convert into absolute magnitudes , we adopt a distance modulus of 18.5 derived by panagia _
et al . _ ( 1991 ) from sn1987a .
a re - evaluation of the cepheid distance calibration , using hipparcos parallaxes , suggests a slight upward revision in the lmc distance modulus to @xmath12 ( madore & freedman 1997 ) , however , our conclusions are insensitive to errors of this order in the distance modulus .
schwering & israel ( 1991 ) determine a foreground reddening of @xmath13 towards field 1 and variations less than 0.02 in @xmath14 between the three fields . allowing for a small amount of internal extinction
, we adopt @xmath15 with a corresponding extinction of @xmath16 . to estimate completeness , a set of artificial stars tests
was performed . at a series of different brightnesses ,
artificial stars were added to each frame in an equally spaced grid and the frames were run through the photometry routine described above .
the grid spacing was chosen so that artificial stars did not add significantly to crowding in the field ; 121 stars were placed on the pc and 529 stars on each of the wfs . the resulting photometry list was compared to the input list and the completeness level was determined as described by holtzman _
et al . _ ( 1997 ) .
we estimate the 90% completeness level to be at @xmath17 in fields 1 and 2 , and @xmath18 in field 3 due to the lower exposure time .
we restrict our analysis to stars brighter than these limits .
the fraction of detected artificial stars and their associated errors were tabulated as a function of magnitude and are used in @xmath19 to simulate observed stellar populations .
these errors include systematic errors due to crowding and random errors from poisson statistics .
observational errors for our simulations ( discussed below ) are determined by randomly sampling from these error distributions .
the stellar density in these regions of the lmc is low and the fields are not crowded ( see holtzman _
et al . _
1997 for an image of field 1 ) . for stars brighter than @xmath20 ,
our observations are not representative due to saturation and the small wfpc2 field of view . in @xmath19
, we correct for small number statistics at the brightest magnitudes using ground based data taken with the mount stromlo and siding spring observatories 1 m telescope ( stappers _ et al .
_ 1997 ) .
three fields adjacent to and including field 1 were observed in v and i. the exposure time for each field was 1000s in each filter .
the total area on the sky , excluding a region around the cluster ngc 1866 and a second smaller cluster , was 622 arcmin@xmath21 .
reduction and photometry were done using standard iraf tasks .
these data are complete to an apparent magnitude of @xmath22 .
photometric consistency between the ground - based and wfpc2 data was checked by comparing 37 stars common to both ; differences in v and i were @xmath23 magnitudes .
color - magnitude diagrams ( cmds ) for the three wfpc2 fields are shown in figure 2 .
the faintest main sequence turnoff for the three fields occurs at @xmath24 , and a clear main sequence extends roughly four magnitudes fainter .
the number of stars in each field is roughly comparable .
error bars plotted in figure 2 are average one @xmath25 errors as determined by the aperture photometry routine .
larger errors in field 3 are a result of the shorter exposure time .
major features in these cmds , such as the main sequence , the main sequence turnoff , and the red giant branch , occur at the same magnitude and color , suggesting similar stellar populations .
we first compare stellar distributions across the lower main sequence as these are sensitive to metallicity variations between the three populations .
we then compare the distributions across the upper main sequence as these probe variations between the recent star formation histories of the three fields . as a statistical test of comparison
, we use the one dimensional kolmogorov - smirnov ( ks ) test .
this test is used to compare the three fields , as well as to compare the observed luminosity functions to simulated stellar populations .
the ks test gives the probability ( p ) that the deviations between two distributions are the same as would be observed if they were drawn from the same population .
two distributions are considered different if the probability that they are drawn from the same parent distribution can be ruled out at a confidence level greater than 95% ( @xmath26 ) .
if two distributions can not be proved different , we infer that the populations are similiar , although the ks test does not imply that these distributions are the same .
we estimate the sensitivity of this test to minor differences in the star formation history using simple simulations described below . as seen in figure 2
, stars appear to be concentrated towards the blue side of the lower main sequence .
the lower main sequence of field 1 appears the most concentrated towards the blue , field 3 appears the least concentrated . for field 3 , this is likely the result of a lower exposure time , but for field 2 may reflect a real difference with the stellar population in field 1 .
the histograms of figure 3 plot the color distributions for fields 1 and 2 in several magnitude bins across the lower main sequence .
the total number of stars in each histogram has been normalized to the number in field 1 .
we used the ks test to statistically compare the distribution of stars across the lower main sequence in these fields . in figure 3 ,
the relatively high values of the ks probability , p= 0.2 , 0.3 and 0.9 , indicate that we can not demonstrate that the two samples are drawn from different populations .
the signal - to - noise in field 3 is lower due to a shorter exposure time . in order to compare lower main sequence distributions
, we add gaussian noise to the higher signal - to - noise field 1 using the average errors from field 3 and perform the ks test . in figure 4 , the distribution of lower main sequence stars in field 3 is compared to the field 1 plus noise distribution .
again , we can not prove that the distributions are different at a high confidence level .
the application of the ks test to the lower main sequence color distributions is weakened by the small number of stars in each comparison , however , it does suggest similar stellar populations in the three field .
although the distribution of stars will also be influenced by age variations and the presence of binaries , the similarity in the mean colors of the lower main sequences suggest that the three fields have similar mean metallicities . to estimate the sensitivity of our tests to differences in metallicity , we use stellar models described below to simulate stellar populations with identical star formation histories but different metallicity distributions .
we find that the ks test is sensitive to metallicity differences if , between the two populations , at least 25% of the stars have a factor of four difference in metallicity .
this fraction decreases to 20% of stars if the metallicity differs by a factor of ten .
these results are robust for several different assumed star formation histories . from the width of the lower main sequence
, we can rule out the possibility that the observed stellar populations have a single metallicity .
the standard deviations of the observed lower main sequence distributions shown in figure 3 are @xmath27 for the three magnitude bins .
field 3 is not included due to the higher noise .
we compare this width with that of a simulated single metallicity population , assuming a constant star formation rate from 12 gyr to the present and a 50% binary fraction .
observational errors are simulated by randomly sampling the error distributions determined from artificial star tests , and include both systematic and random errors .
independent of assumed metallicity , the standard deviation of the lower main sequence was @xmath28 for the same magnitude bins .
we find that the difference between the observed and model variances is significant based on a f - test .
populations having a smaller age range or a lower assumed binary fraction will have narrower distributions . therefore , the observed lower main sequence is wider than is expected for a single metallicity population . since this is a differential comparison , we expect it to hold regardless of the adopted stellar models . the derivation of the absolute metallicity is discussed in @xmath29 and may be model dependent .
the distribution of stars across the upper main sequence is shown in figure 5 .
the width of the main sequence is broader than would be expected from a single age population . from stellar evolution models , stars brighter than @xmath30 evolve steadily redward during their main sequence lifetimes .
thus , a uniform distribution across the upper main sequence suggests no intense star formation bursts have occurred during the lifetimes of these more massive stars .
_ ( 1996 ) show that the distribution of upper main sequence stars in field 1 is roughly consistent with a star formation rate constant for the past 3 gyr .
the ks test gives no evidence that the distributions of stars across the upper main sequence in the three fields are different .
this test is not sensitive to variations in the star formation history more recent than 0.1 gyr , due to small number statistics at the brightest magnitudes . a second distinct main sequence turnoff discussed by gallagher _
et al . _ ( 1996 ) and seen near @xmath31 in figure 2a is not observed in the remaining two fields .
_ interpret this turnoff and associated excess of stars at @xmath32 in the hertzsprung gap as signatures of a short star formation burst occurring @xmath33 gyr ago .
although many stars populate this region of the cmds in field 2 and 3 , the lack of a single distinct subgiant branch in these fields argues against a short ( @xmath34 gyr ) , global 2 gyr burst throughout the lmc .
the subgiant excess observed in field 1 may arise from a statistical fluke , the remnants of a localized star formation burst , or a dissolved cluster .
the observed , uncorrected , differential luminosity functions are shown in figure 6 .
we perform the ks test between fields 1 and 2 over the range @xmath35 and between field 1 and field 3 over the range @xmath36 .
the resulting ks probabilities between fields are @xmath37 and @xmath38 , thus we are not able to show that the three luminosity functions are different .
we conclude , from analysis of the luminosity function and distribution of stars across the main sequence , that the three observed regions in the lmc field contain statistically indistinguishable stellar populations .
in addition , we have calculated r - ratios as defined by bertelli _
_ ( 1992 ) and discussed below ( @xmath19 ) , which compare the number of stars in different evolutionary phases ; these are shown in table 2 .
these ratios are also the same between the three fields , within the errors determined by number statistics .
we can not prove that the star formation histories in the three fields are different ; however , this does not imply that they are identical . to estimate the sensitivity of our tests to variations in star formation history , we simulate a simple star formation history with a constant star formation rate .
we compare this to models in which star formation is turned off completely for short lengths of time at different epochs .
we find that our statistical tests are unable to conclusively distinguish between such models for variations in star formation rate over periods shorter than 1 gyr anytime in the past 4 gyr , or over periods shorter than 2 gyr anytime before 4 gyr ago .
we next compare our observations to stellar models using similar statistical tests in order to place constraints on possible star formation histories .
as we have shown that our methods can not distinguish between the three stellar populations , we combine the observations of the three fields to improve number statistics .
to determine the star formation history in the three fields , we compare our observations to simulations made using the stellar evolution models published by the padua group ( bertelli _ et al . _
1994 and references therein ) .
these isochrones range in metallicity from @xmath39 ( -1.7@xmath40\leq$]0.4 ) and are calculated for stellar masses down to @xmath41 .
depending on the metallicity , this corresponds to an absolute magnitude of @xmath42 , roughly the magnitude limit of our observations .
the padua models are calculated with mild convective overshoot and the most recent livermore group radiative opacities ( iglesias _ et al . _
ubvri magnitudes for these models have been calculated by bertelli _
et al . _ ( 1994 ) .
there is evidence for nonsolar abundance ratios in the lmc , such that the @xmath0-element are enhanced relative to the solar ratio ( luck & lambert 1992 ) .
although @xmath0-enhanced models are not currently available , salaris _ et al . _ ( 1993 ) have found that under some conditions , @xmath0-enhanced isochrones are well mimicked by scaled solar metallicity isochrones .
the effect of @xmath0-element enhancement is to shift a solar abundance isochrone towards the red , which would lead to an overestimate of our derived metallicities . comparing the observed cmds to single age isochrones
, we find the blue edge of the upper main sequence is best matched by an isochrone of metallicity z=0.008 , whereas the red giant branch can be fit by either an old to intermediate age , metal poor isochrone ( @xmath43 gyr and z=0.0004 ) or a young , higher metallicity isochrone ( @xmath44 gyr and z=0.001 ) .
this point is well illustrated in figure 4 of holtzman _
et al . _ ( 1997 ) .
problems with the stellar atmospheres at lower temperatures and/or the evolutionary models of giant stars may be responsible for the apparent mismatch .
alternatively , it may be related to our use of solar abundance ratio isochrones .
because of these possible problems , we do not use the color of the giant branch to derive stellar population parameters .
however , some of the derived parameters , in particular metallicities , are sensitive to the model colors of main sequence stars .
we note that our constraints on such parameters are derived assuming that these stellar models are perfect ; we allow for random errors in the observations but not for systematic errors in the models .
the presence of a bright main sequence , seen in figure 2 , suggests recent star formation activity , whereas the faintest main sequence turnoff at @xmath24 implies an older star formation epoch .
we therefore simulate cmds for a mixture of simple stellar populations . throughout the simulations
, we assume that 50% of the stars are binaries with uncorrelated masses ( reid 1991 ) . in order to preserve isochrone shape during age interpolation , each isochrone is resampled into 100 equally spaced mass points within each of 9 evolutionary epochs .
no interpolation is made in metallicity .
the mass and age of each star are chosen according to an input initial mass function and star formation history .
the absolute magnitude and color of each star can be determined for any desired age by interpolating point by point over the resampled isochrones .
an apparent magnitude is determined based on the extinction and distance modulus .
the star is considered detected or rejected according to the completeness histograms calculated during the artificial star tests described above .
an observational error is given to each detected star by randomly sampling the error distribution appropriate for the star s magnitude determined from the artificial star tests ( see @xmath45 ) .
these errors include both random and systematic effects .
the number of free parameters in determining a star formation history is large .
our observations do not provide enough constraints to justify an exhaustive search of parameter space .
instead , we choose to discuss six representative formation histories : constant star formation throughout the history of the lmc , the two preferred histories of bertelli _ et al . _ ( 1992 ) and vallenari _ et al . _ 1996b , two proposed histories of holtzman
et al . _ ( 1997 ) , and a formation history motivated by the observed age distribution of lmc globular clusters . in all cases , we assume the age of the oldest lmc stars to be 12 gyr based on age estimates of the oldest lmc globular clusters ( van den bergh 1991 ) . the parameters used in each simulation are shown graphically in figure 7 .
to compare observed and theoretical stellar populations , we use a combination of two methods .
first , we compare observed and model luminosity functions using the 1-d ks test as described in @xmath46 .
the three fields are combined to create the observed luminosity function and are compared with models over a range @xmath47 .
a star formation history is considered acceptable if the probability , p , that the luminosity function is drawn from the same population is greater than 5% .
the sensitivity of this test to variation in the star formation history is the same as that discussed in @xmath48 . in comparing luminosity functions without regard to color
, some star formation history information is lost , especially for stars brighter than the main sequence turnoff .
thus , in addition to luminosity function fitting , we use the r - method described in detail by bertelli _
et al . _ ( 1992 ) for @xmath49 .
briefly , this method defines three stellar number ratios , each sensitive to different parameters in the star formation history .
the first ratio is defined as : @xmath50 the separation between main sequence and red giant stars is determined by the lines shown in figure 2 , and is consistent throughout the analysis .
the next two ratios compare the number of bright to faint stars on the main sequence and the red giant branch .
the magnitude separating the upper and lower regions is defined at @xmath51 , chosen to be below the red clump .
the ratios are defined as : @xmath52 @xmath53 the r - ratios depend , in different ways , on the slope of the initial mass function and the relative number of young and old stars ; see bertelli _
et al . _ ( 1992 ) for a more extended discussion of these dependencies .
the r - ratios for the three observed fields , as well as the ratios resulting from combining the three fields , are shown in table 2 .
the errors in this table are one @xmath25 errors as determined from number statistics .
the wfpc2 fields do not provide representative numbers of stars at the brightest magnitudes due to saturation and the small field of view .
we corrected for this by combining the _ hst _ data with ground - based data covering a significantly larger field and including field 1
. a more detailed analysis of these data was presented by stappers _
et al . _ ( 1997 ) . to combine star counts from ground and space - based data directly , we applied a scale factor determined by the ratio of observed areas .
we use ground - based star counts for magnitudes brighter than @xmath51 to recalculate the r - ratios of the combined field . as compared to the uncorrected ratios ( table 2 , second to last row ) , the use of ground - based data results in an average 10% corrections to the r - ratios ( table 2 , last row ) .
constraints from the r - ratios , particularly @xmath54 , may be less secure than those based on the main sequence because of the larger uncertainties in modelling later phases of stellar evolution .
we compare the observed luminosity function and r - ratios with those calculated for the six star formation histories shown in figure 7 .
the observed and computed r - ratios , as well as the ks probability resulting from a comparison of the observed and model luminosity functions , are given in table 3 .
the observed and simulated luminosity functions for each model are shown in figure 8 .
the simulated luminosity functions are normalized to match the observations at @xmath55 .
the simplest star formation history tested assumes a salpeter imf and a constant star formation rate since the formation of the lmc 12 gyr ago ( figure 7a ) . in examining the parameter @xmath56
, this simple formation scenario does not produce enough bright main sequence stars by more than a factor of two relative to the number of observed evolved stars .
this deficiency has motivated most modelers to include an enhancement in the recent star formation rate .
bertelli _ et al .
_ ( 1992 ) observed a field @xmath57 southwest of field 1 . using the r - method
, they derive a star formation history in which the star formation rate was initially low and then increased by a factor of ten 4 gyr ago , as shown in figure 7b .
although this model reproduces the observed r - ratios reasonably well , it does not match the observed luminosity function .
as shown in figure 8b , this formation scenario produces too many bright ( @xmath58 ) stars relative to faint stars .
we note that the 1992 padua stellar models used to derive this history allow for more convective overshoot than the 1994 models used in this paper .
the use of more recent models decreases the inferred time when the star formation increase rate began to approximately 2 gyr ago ( bertelli , private communication ) .
such a star formation history was used by vallenari _
_ ( 1996b ) to match observations in several of their fields and its luminosity function is shown in figure 8c .
it produces even more bright stars relative to the number of observed faint stars , and is ruled out by our observed luminosity function . in their simulations ,
et al . _ allow for interpolation in metallicity , whereas our models use discrete metallicity isochrones .
we find that the luminosity functions of the three individual metallicities which contribute to the final formation scenario are inconsistent with the observations in the same direction as the composite vallenari _
_ luminosity function .
therefore , this simplification is not the source of discrepancy between the vallenari _
_ scenario and our observations .
holtzman _ et al .
_ ( 1997 ) find that a steeper imf slope is necessary in order for a 4 gyr , ten - fold star formation rate increase to match the observed luminosity function .
for an imf slope @xmath59 , this star formation history ( figure 7d ) is consistent with the observed luminosity function , but not with the r - ratios observed in our fields . of published star formation histories for the lmc field ,
the only one which reproduces both the luminosity function and r - ratios of our observation is that of holtzman _
et al . _ ( 1997 ) .
in this formation scenario , the star formation rate remains constant for a majority of the lmc history and is increased by a factor of three from 2 gyr ago to the present ; the star formation rate is also slightly higher for the oldest stars ( figure 7e ) . in contrast to the previous three scenarios which produce primarily young stellar populations , this star formation history produces a population with roughly equal number of stars older and younger than 4 gyr .
the simulated cmd for this star formation history is shown in the right panel of figure 9 .
a star formation history based on the age distribution of lmc globular clusters can not fit the observed field population .
the age distribution of lmc clusters is bimodal ( figure 7e ) with @xmath60 of clusters formed between @xmath61 years and 3 gyr ago , and @xmath62 of clusters having ages between 10 and 12 gyr ( van den bergh 1991 ) .
there are almost no known intermediate age ( 3 - 10 gyr ) clusters in the lmc ( girardi _ et al . _ 1995
; however see sarajedini _ et al . _
1995 ) , however , an intermediate population is necessary to reproduce our observations .
if no clusters have been destroyed , we conclude that the star formation history of lmc globular clusters is not mimicked by the field population .
it is possible to predict the chemical history of the lmc from its star formation history using the simple closed box model of chemical evolution .
we assume a one - zone evolution model with no infall or outflow , zero initial metal content and instantaneous recycling ( searle & sargent 1972 ) . this model has successfully predicted the relationship between metallicity and current gas fraction in magellanic irregular galaxies , although it has less success predicting this relationship in larger spiral systems ( binney & tremaine 1987 ) .
we assume a present day gas to total lmc mass ratio of @xmath63 ( cohen _ et al . _
1988 ) and an effective yield of @xmath64 , chosen so that the present day metallicity matches that inferred from the upper main sequence ( @xmath65 ) .
we compare the predicted chemical evolution for two star formation histories in figure 10 . in the upper panel of figure 10 , the holtzman _
et al . _ ( 1997 ) star formation history suggests that the metallicity in the lmc has doubled in the past 2 gyr .
the vallenari _
et al . _ ( 1996 ) formation history ( bottom panel ) implies a factor of five metallicity increase in the past 2 gyr . the chemical evolution predicted by the holtzman _
et al . _
model is consistent with planetary nebula observations by dopita _
et al . _ ( 1997 ) which suggest the metallicity of the lmc has almost doubled in the last 2 gyr .
we also note that a significant metal poor population is predicted by the closed box model , regardless of the details of the star formation history . for an effective yield of @xmath64 ,
the closed box model predicts 22% of stars have metallicities less than z=0.001 .
this fraction is inversely proportional to the assumed yield .
simulated lower main sequences suggest a similar fraction of lmc field stars are metal poor .
lower main sequence cross sections are shown for two star formation histories in figure 11 . in the left column ,
vallenari _ et al .
_ s model assumes a metallicity range z=0.008 - 0.001 .
these distributions are displaced to the red of the observations by as much as a tenth of a magnitude and are significantly narrower .
although some redward evolution occurs for low mass stars during their main sequence lifetime , changing the star formation history alone is not enough to explain this color shift .
the holtzman _
et al . _
model is better able to fit the lower main sequence , as it includes an old , metal poor component as shown in the right column of figure 11 . in this simulated population ,
20% of stars have a metallicity z=0.0004 ( [ fe / h]= -1.7 ) .
none of the models considered here , however , include chemical evolution in a fully self - consistent manner .
direct measurements of metallicity in the lmc field have been limited to bright stars , but possibly suggest a similar fraction of metal poor stars . olszewski ( 1993 ) spectroscopically determined metallicities for 36 red giant stars in an outer lmc field near ngc 2257 and found that 8 or 9 ( @xmath66 ) of these stars had metallicities below z=0.001 ( [ fe / h]=
more metallicity observations are needed to determine the size of a metal poor component in the lmc field ( suntzeff 1997 ) .
we present deep wfpc2 observations of three fields in the outer disk lmc . we find no conclusive evidence for variation in the stellar populations between the three fields based on the morphologies of the color - magnitude diagrams , the luminosity functions , and the relative numbers of stars in different evolutionary stages . in apparent contrast to our results , vallenari _ et al .
_ ( 1996b ) find significant variations in the star formation history correlated with azimuthal angle in the lmc field . a direct comparison with their results , however , is difficult . the r - ratios of our field 1 agree with those calculated for the nearly overlapping vallenari _
et al . _
ngc 1866 field .
field 3 is located reasonably close to field 1 , therefore the similarity of this field to field 1 provides no direct contradiction with the vallenari _
_ results .
a direct discrepancy comes from field 2 , which is located @xmath67 from the vallenari _
_ field lmc-61 .
we find significant disagreement in the ratio @xmath68 between these two fields .
this discrepancy may be due to the small difference in location or to some systematic error in one or both of the samples . a much larger survey is required to determine whether a correlation exits between the star formation history and position in the lmc field ( zaritsky , harris & thompson 1997 ) .
other evidence that the star formation history varies within the lmc comes from elson _
et al . _ ( 1997 ) , who present evidence that the stellar population in the lmc bar is different from those presented in this paper .
they analyze _ hst _ observations for a field in the bar of the lmc and identify additional peaks in the color distribution between @xmath69 not associated with the red giant branch or the most recent epoch of star formation .
they attribute these peaks to a burst of star formation between 1 and 2 gyr depending on the assumed metallicity .
we do not find evidence for this population in our observed color distribution .
_ associate this population with the formation of the lmc bar .
we have compared our observations with stellar models to place constraints on possible star formation histories in the three fields .
these constraints are an improvement over previous results as they incorporate both _
hst _ and ground - based data , allowing measurements of the deep main sequence luminosity function , the distribution of stars in the upper main sequence band , and the relative number of bright stars which probe different evolutionary phases . of previously considered star formation histories ,
the only one which is consistent with all of our observations has a star formation rate which is roughly constant for 10 gyr , then increases by a factor of three for the past 2 gyr .
contrary to many previous models , this produces a population which is _ not _ dominated by young stars .
although the star formation history of the lmc is clearly more complicated , this simple picture should provide a useful guide to understanding the formation of our nearest neighbor .
bertelli , g. , bressan , a. , chiosi , c. , fagotto , f. & nasi , e. 1994 , a&as , 106,275 bertelli , g. , mateo , m. , chiosi , c. , & bressan , a. 1992 , apj , 388 , 400 binney , j. & tremaine , s. 1987 , _ galactic dynamics _ , ( princeton university press : princeton , n.j . ) cohen , r. s. _ et al .
_ apj 331 , l95 dopita , m. _ et al . _ 1997 ,
apj , 474 , 188 elson , r. , gilmore , g. & santiago , b. 1997 , mnras , 289 , 157 gallagher , j. s. , _ et al . _ 1996 ,
apj , 466 , 732 girardi , l. _ et al . _ 1995 , a&a , 298 , 87 iglesias , c. a. , rogers , f. j. , & wilson , b. g. 1992 , apj , 397 , 717 holtzman , j. a. _ et al .
_ 1995a , pasp , 107 , 156 holtzman , j. a. _ et al . _ 1995b , pasp , 107 , 1065 holtzman , j. a. _ et al . _
1997 , aj , 113 , 656 luck , r. e. & lambert , d. l. 1992 , apjs , 79 , 303 madore , b. & freedman , w. 1997 apjl , in press olszewski , e.w .
1993 , in asp conf .
48 , the globular cluster - galaxy connection , ed .
g. smith & j. brodie ( san francisco : asp ) , 351 panagia _
et al . _
1991 , apj , 380 , l23 .
et al . _
1995 , a&a , 304 , 69 .
reid , n. 1991 aj , 102 , 1428 .
sandage , a. 1961 , _ the hubble atlas of galaxies _ , ( carnegie institution of washington : washington , d.c . ) sarajedini , a. , lee , y. , lee , d. 1995 apj 450 , 712 schwering , p. g. , & israel , f. p. 1991 , a&a , 246 , 231 searle , l. , & sargent , w. l. w. 1972 , apj , 173 , 25 salaris , m. , chieffi , a. , & straniero , o. 1993 , apj , 414 , 580 stappers , b. j. , _ et al . _ 1997 ,
pasp , 109 , 292 stryker , l. , 1984 , apjs , 55 , 127 suntzeff , n. 1997 , bull .
a. a. s. , 190 , 3501 vallenari , a. , chiosi , c. bertelli , g. , & ortolani , s. 1996a , a&a , 309 , 358 vallenari , a. , chiosi , c. bertelli , g. , aparicio , a. , & ortolani , s. 1996b , a&a , 309 , 367 van den bergh , s. 1991 , apj , 369 , 1 westerlund , b. e. , linde , p. & lynga , g. 1995 , a&a , 298 , 39 zaritsky , d. harris , j. , & thompson , i. 1997 , astro - ph/9709055 field 1 & @xmath72 & @xmath73 & @xmath74 + field 2 & @xmath75 & @xmath76 & @xmath77 + field 3 & @xmath78 & @xmath79 & @xmath80 + combined fields & @xmath81 & @xmath82 & @xmath83 + combined + ground & @xmath84 & @xmath85 & @xmath86 + c c c c c c + + & & & & & + observed values & @xmath84 & @xmath85 & @xmath87 & + constant star & @xmath88 & @xmath89 & @xmath90 & @xmath91 + formation & & & & + bertelli _ et al .
_ & @xmath92 & @xmath93 & @xmath94 & @xmath95 + ( 1992 ) & & & & + vallenari _ et al . _ & @xmath96 & @xmath97 & @xmath98 & @xmath99 + ( 1996b ) & & & & + holtzman _ et al . _ & @xmath100 & @xmath101 & @xmath102 & 0.14 + ( 1997 ) @xmath59 & & & & + holtzman _ et al
. _ & @xmath103 & @xmath104 & @xmath105 & 0.17 + ( 1997 ) @xmath106 & & & & + lmc cluster & @xmath107 & @xmath108 & @xmath109 & @xmath99 + age distribution & & & & + | we present _ hst _ photometry for three fields in the outer disk of the lmc extending approximately four magnitudes below the faintest main sequence turnoff .
we can not detect any strongly significant differences in the stellar populations of the three fields based on the morphologies of the color - magnitude diagrams , the luminosity functions , and the relative numbers of stars in different evolutionary stages .
our observations therefore suggest similar star formation histories in these regions , although some variations are certainly allowed .
the fields are located in two regions of the lmc : one is in the north - east field and two are located in the north - west . under the assumption of a common star formation history ,
we combine the three fields with ground - based data at the same location as one of the fields to improve statistics for the brightest stars .
we compare this stellar population with those predicted from several simple star formation histories suggested in the literature , using a combination of the r - method of bertelli _ et al .
_ ( 1992 ) and comparisons with the observed luminosity function . the only model which we consider that is not rejected by the observations is one in which the star formation rate is roughly constant for most of the lmc s history and then increases by a factor of three about 2 gyr ago .
such a model has roughly equal numbers of stars older and younger than 4 gyr , and thus is not dominated by young stars .
this star formation history , combined with a closed box chemical evolution model , is consistent with observations that the metallicity of the lmc has doubled in the past 2 gyr . |
a companion paper @xcite reviews magnetic bunch compression and shows a schematic of our system .
i assume the reader is familiar with this paper .
the aps bunch compressor design is an outgrowth of studies @xcite by p. emma and v. bharadwaj of stanford linear accelerator center ( slac ) .
they explored a number of designs , including symmetric and asymmetric four - dipole chicanes .
starting from this work , i investigated a large number of configurations with various values of @xmath1 , asymmetry , and final current . for each configuration ,
detailed longitudinal and transverse matching was performed , followed by tracking with csr and wakefields .
then , sensitivity analysis was performed for all configurations , followed by jitter simulations for the least sensitive configurations .
this work relied on elegant @xcite , a 6-d code with a fast simulation of csr effects , plus longitudinal and transverse wakefields .
elegant also performs optimization of actual tracking results , such as bunch length , energy spread , and emittance .
simulation of the linac uses the rfca element , a matrix - based rf cavity element with exact phase dependence .
our linac has quadrupoles around the accelerating structures .
hence , i split each 3-m section into about 20 pieces , between which are inserted thin - lens , 2nd - order quadrupole elements .
a series of such elements is used for each quadrupole .
a green s function technique is used to model wakefields , using a tabulation of the slac - structure wake functions provided by p. emma @xcite . to reduce running time ,
one longitudinal wake element is used per 3-m section , which is a good approximation for relativistic particles . for transverse wakes ,
i used one wake element per rf cavity element ( about 20 per section ) . the csr model used by elegant
is based on an equation @xcite for the energy change of an arbitrary line charge distribution as a function of the position in the bunch and in a bending magnet .
details of this model will be presented by the author at an upcoming conference .
effects of changes in the longitudinal distribution within a dipole are included .
csr in drift spaces is included by propagating the terminal csr `` wake '' in each bend through the drifts with the beam .
longitudinal and transverse matching has the goal of providing configurations for the 300-a and 600-a leutl operating points @xcite .
the starting point for the simulations is macro particle data generated @xcite with parmela , giving the 6-d distribution after the photoinjector ( pi ) .
see figure 1 in @xcite .
longitudinal matching involves adjusting the phase and voltage of l2 ( see @xcite for nomenclature ) to obtain the desired current and energy after the chicane .
then , l4 and l5 are adjusted to minimize the energy spread and obtain the desired final energy .
longitudinal matching includes longitudinal wakefields , rf curvature , and higher - order effects in the beam transport , by matching _
tracked _ properties of the simulated beam .
figure [ fig : longitdist ] shows the longitudinal phase space for the 300-a case with @xmath2 mm , which exhibits a current spike of nearly 1200 a. the matching ignores this spike ( which is shorter than a slippage length for 530 nm ) because of the way `` current '' is defined @xcite . following longitudinal matching , transverse matching
is done for each configuration .
initial twiss parameters are obtained from the rms properties of the parmela beam .
starting values for the quadrupoles were obtained from matching `` by hand '' for one configuration .
four sequential elegant runs work the beta functions down the linac .
the most important constraints maintain small beta functions in the linac ( for transverse wakefield control ) , small horizontal beta in dipole b4 ( to reduce csr effects ) , and matching for the emittance measurement sections .
figure [ fig : twiss ] shows sample twiss parameters in the chicane region .
the matching is highly automated , so that only the desired beam current and energy needs to be specified .
evaluation of tolerances and randomized simulations are also automated , being set up by scripts from the corresponding matching runs .
transfer of data between simulation stages is handled using sdds files and scripts @xcite , reducing errors and increasing the number of configurations that can be examined .
for example , a script is used to scan all configurations and give power supply specifications . a distributed queue
utilizing 50 workstations is used to run the simulations .
figure [ fig : emittrends ] shows emittance vs. @xmath1 for the symmetric ( a=1 ) and asymmetric ( a=2 ) cases at 300 a and 600 a. for 300 a , the symmetric and asymmetric cases are very similar . for 600 a , the difference is 10% or more , which should be measurable .
one surprise in figure [ fig : emittrends ] is that the emittance does not uniformly increase as @xmath3 increases , even though elegant shows the expected monotonic increase ( due to csr ) vs bending angle for a single dipole with a constant input beam distribution .
this is apparently due to variation in the compressed bunch distribution between cases with the same `` current '' but different @xmath1 . for smaller @xmath3 ,
there are higher current spikes at the head of the bunch , leading to a larger and more rapidly changing csr wake , which in turn leads to larger emittance growth .
the effect is even more pronounced in the 1200-a cases ( not shown ) .
insertion of the scraper between b2 and b3 to remove the low - energy part of the beam can reduce the height and width of the current spike , resulting in lower emittance .
unfortunately , this also reduces the current in the rest of the bunch considerably .
earlier simulations showed that emittance trends can be changed significantly by inconsistent values of the horizontal beta function at the exit of b4 .
all of these subtleties will make for difficult interpretation of experiments in which @xmath1 is varied .
however , because compression to different currents for fixed @xmath1 involves only adjustment of the rf phases and voltages , comparision of the emittance growth for different amounts of compression should be more straightforward .
tolerances are driven by the fel gain length , trajectory , and wavelength stability requirements @xcite .
the 10% rms gain length variation limit is easy to use in elegant as it computes fel performance directly using xie s parameterization @xcite .
beam trajectory limits ( @xmath450 @xmath0 m , @xmath450@xmath0r ) are included separately as they are not incorporated into xie s formula .
the 1-nm rms wavelength variation limit is a challenging goal at 530 nm as it puts a 0.1% limit on energy variation .
the analysis begins by running single - parameter `` sweeps '' to assess the effect on the constrained quantities ( gain length , trajectory , and wavelength ) of accelerator parameters ( e.g. , rf phase ) .
sweeps included rf phase and voltage ; photoinjector timing , charge , and energy ; and chicane dipole strength . from these sweeps ,
a script determines the limit on each parameter change due to the various specifications , showing that configurations with the largest @xmath1 are least sensitive to difficult - to - control timing and phase errors .
these configurations experience the most emittance degradation from csr , but tend to yield the shortest gain length as they have the smallest energy spread ( l2 being closer to crest ) .
.selected sweep limits for @xmath2 mm [ cols="<,^,^",options="header " , ]
the technical note @xcite by p. emma and v. bharadwaj provided a valuable starting point .
i acknowledge helpful discussions and assistance from h. friedsam , e. lessner , j. lewellen , s. milton , and g. travish .
j. lewellen provided the pi beam distribution data . | a magnetic bunch compressor was designed and is being commissioned to provide higher peak current for the advanced photon source s ( aps ) low - energy undulator test line ( leutl ) free - electron laser ( fel ) @xcite
. of great concern is limiting emittance growth due to coherent synchrotron radiation ( csr ) .
tolerances must also be carefully evaluated to find stable operating conditions and ensure that the system can meet operational goals .
automated matching and tolerance simulations allowed consideration of numerous configurations , pinpointing those with reduced error sensitivity .
simulations indicate tolerable emittance growth up to 600 a peak current , for which the normalized emittance will increase from 5 to about 6.8 @xmath0 m .
the simulations also provide predictions of emittance variation with chicane parameters , which we hope to verify experimentally . |
the study of lepton flavor violating ( lfv ) processes offers a good opportunity for shedding light on the possible physics beyond the standard model ( sm ) .
muon - to - electron ( @xmath0 ) conversion in nuclei @xmath2 is commonly recognized as one of the most sensitive probes of lepton flavor violation and of the related physics behind it ( for reviews , see @xcite ) . at present , on the experimental side there is one running conversion experiment , sindrum ii @xcite , and two planned ones , meco @xcite and prime @xcite .
so far this lfv process has not been observed and experimental results correspond to the upper limits on the conversion branching ratio @xmath3 the current and expected limits from the above mentioned experiments are presented in table i. as is known from previous studies ( see , for instance , ref .
@xcite and references therein ) and will be discussed later in the present paper , these experimental bounds allow to set stringent limits on the mechanisms of conversion and the underlying theories of lfv .
the theoretical studies of conversion , presented in the literature , cover various aspects of this lfv process , elaborating adequate treatment of the structure effects @xcite of the nucleus participating in the reaction and , considering underlying mechanisms of lfv at the level of quarks within different scenarios of physics beyond the sm ( see @xcite and references therein ) . in general the conversion mechanisms
can be classified as photonic and non - photonic . in the former case
the conversion is mediated by the photon exchange between the lfv leptonic vertex and the ordinary electromagnetic nuclear vertex .
the non - photonic mechanisms are based on the 4-fermion contact lepton - quark lfv interactions .
these two categories of mechanisms differ significantly from each other since they receive different contributions from the new physics and require different treatment of the effects of the nucleon and the nuclear structure . in the present paper
we continue studying the non - photonic meson exchange mechanism of conversion .
previously @xcite we analyzed the vector - meson mediation of the conversion .
the contribution of this mechanism to the coherent conversion results in some important issues for the physics beyond the sm absent in the case of the conventional direct lepton - nucleon interaction . here ,
we extend our analysis to the scalar - meson exchange mechanism , which completes the study of meson exchange contributions to the coherent mode of the conversion .
the effective lagrangian @xmath4 describing the coherent conversion at the quark level can be written in the form @xcite @xmath5 , \label{eff - q}\end{aligned}\ ] ] where the lepton and quark currents are defined as : @xmath6 in eqs .
( [ eff - q ] ) and ( [ lepton - currents ] ) the summation runs over all the quark species @xmath7 .
the parameter @xmath8 with the dimension of mass is the characteristic high energy scale of lepton flavor violation attributed to new physics .
the dimensionless lfv parameters @xmath9 in eq .
( [ eff - q ] ) depend on a concrete lfv model . in the present study
we treat these parameters as phenomenological parameters to be constrained from the experiment .
the quark level lagrangian ( [ eff - q ] ) generates the effective lepton - nucleon interactions that can be specified in terms of an effective lagrangian on the nucleon level @xmath10\ , .
\label{eff - n } \end{aligned}\ ] ] here , the isoscalar @xmath11 and isovector @xmath12 nucleon currents are @xmath13 and @xmath14 where @xmath15 is the nucleon isospin doublet , @xmath16 and @xmath17 .
the summation over the double indices @xmath18 and @xmath19 is implied in ( [ eff - n ] ) .
the nucleon lagrangian ( [ eff - n ] ) is the basis for the derivation of the nuclear transition operators .
naturally , the lagrangian in terms of effective nucleon fields ( [ eff - n ] ) is equivalent to the quark level lagrangian ( [ eff - q ] ) . the former lagrangian is supposed to appear after the hadronization from the quark level lagrangian ( [ eff - q ] ) and , therefore , must correspond to the same order @xmath20 in inverse powers of the lfv scale . to make
a bridge between the underlying lfv physics and observables one needs to relate the lepton - nucleon lfv parameters @xmath21 in eq .
( [ eff - n ] ) to the lepton - quark lfv parameters @xmath22 in eq .
( [ eff - q ] ) .
this implies a certain hadronization prescription which specifies the way in which the effect of quarks is simulated by hadrons . in the absence of a true theory of hadronization we rely on some reasonable assumptions and models . in refs .
@xcite we considered the two mechanisms of nuclear @xmath23 conversion : direct nucleon mechanism and vector - meson exchange between nucleon and lepton currents .
we found that the vector - meson exchange plays an important role in the coherent muon - electron conversion .
now , we extend this analysis to the scalar sector of the lfv lagrangians in eqs .
( [ eff - q ] ) and ( [ eff - n ] ) . as in the case of the vector currents , considered in refs .
@xcite , here we distinguish the following two hadronization mechanism .
the first one is the direct embedding of the quark currents into the nucleon ( fig.1a ) , which we call direct nucleon mechanism ( dnm ) .
the second mechanism consists of two stages ( fig.1b ) .
first , the quark currents are embedded into the interpolating scalar meson fields which then interact with the nucleon currents .
we call this possibility meson - exchange mechanism ( mem )
. in general one expects all the mechanisms to contribute to the coupling constants @xmath21 in eq .
( [ eff - n ] ) . however , at present the relative amplitudes of different mechanisms are unknown . in view of this problem
we assume for the first approximation , that only one mechanism is operative and estimate its contribution to the process in question .
this allows us to evaluate the importance of a specific mechanism .
let us update the contribution of the direct nucleon mechanism derived in ref .
the relation between the quark - lepton and nucleon - lepton lfv parameters in eqs .
( [ eff - q ] ) and ( [ eff - n ] ) takes in this case the form @xmath24}^{(3 ) } = \frac{1}{2}\eta_{bs}^{(3 ) } ( g_{s}^{u } - g_{s}^{d } ) , \\
\nonumber & & \alpha_{bs[dnm]}^{(0 ) } = \frac{1}{2}\eta_{bs}^{(0 ) } ( g_{s}^{u } + g_{s}^{d } ) + \eta_{bs}^{s } g_{s}^{s},\end{aligned}\ ] ] where @xmath25 and @xmath26 are the isoscalar and isovector quark couplings .
the form factors @xmath27 are related to the scalar condensates in the nucleon @xcite @xmath28 since the maximal momentum transfer @xmath29 in @xmath30 conversion is much smaller than the typical scale of the nucleon structure we can safely neglect the @xmath31-dependence of the nucleon form factors @xmath27 . at @xmath32 these form factors
are related to the corresponding meson - nucleon sigma - terms : @xmath33\ , , \,\,\ , \sigma_{k n}^{i=1 } & = & \frac{\hat m + m_s}{4 } \ , [ g_{s}^{u } - g_{s}^{d } ] \ , , \,\,\
, y_n = \frac{2 g_{s}^{s}}{g_{s}^{u } + g_{s}^{d}},\end{aligned}\ ] ] where @xmath34 and @xmath35 are the masses of current quarks ; @xmath36 is the strangeness of the nucleon . for these parameters we use the following values @xcite in our analysis @xmath37 the canonical value of the @xmath38 sigma term @xmath39
mev @xcite was originally extracted from the dispersional analysis of @xmath38 scattering data taking into account chiral symmetry constraints . in particular ,
the value of the sigma - term , @xmath39 mev , has been deduced from the analysis of two quantities : @xmath40 mev , the scalar nucleon form factor at the cheng - dashen point @xmath41 , and the difference @xmath42 mev @xcite as induced by explicit chiral symmetry breaking .
the value of the isovector kaon - nucleon sigma - term @xmath43 was estimated in ref .
@xcite using the baryon mass formulas : @xmath44 substituting the above values of the hadronic parameters to eq . ( [ sigma ] )
we obtain for the scalar nucleon form factors : @xmath45 we have to point out that these values contain appreciable theoretical and experimental uncertainties , as seen from their derivation ( for the possible error bars see , for instance , ref .
@xcite ) . in our analysis of the dnm contribution
to we take the numbers from eq .
( [ alpha ] ) as central values of the scalar nucleon form factors .
note , in ref .
@xcite the different set of the values for the nucleon scalar form factors was derived on the basis of the qcd sum rules input parameters , which overestimates the pion - nucleon sigma - term and the strangeness of the nucleon .
in the following we turn to the scalar meson - exchange mechanism of conversion .
although the status of scalar mesons is still unclear @xcite we think it is reasonable to study their effect in the conversion since their contribution is associated with the experimentally most interesting coherent mode of this exotic process .
the lightest unflavored scalar mesons are the isoscalar @xmath46 and the isotriplet @xmath47 states .
the former in the context of the nonlinear realization of chiral symmetry can be treated as a resonance in the @xmath48 system ( see detailed discussion , e.g. in refs .
for simplicity we neglect a possible small strange content of the isoscalar meson and treat this state as @xmath49 .
we derive the lfv lepton - meson effective lagrangian in terms of the interpolating @xmath50 and @xmath51 fields . retaining all the interactions consistent with lorentz invariance
, we obtain the general form of this lagrangian : @xmath52 \end{aligned}\ ] ] with the unknown dimensionless coefficients @xmath53 to be determined from the hadronization prescription .
we assume that the lagrangian to be generated by the quark - lepton lagrangian ( [ eff - q ] ) , and , therefore , all its terms have the same suppression @xmath54 with respect to the large lfv scale @xmath8 . another scale in the problem is the hadronic scale @xmath55 gev which adjusts the physical dimensions of the terms in eq .
( [ eff - ls ] ) . in the lagrangian in eq .
( [ eff - ls ] ) we neglect derivative terms since their contribution to conversion is suppressed by a factor @xmath56 . in order to relate the parameters @xmath53 of the lagrangian ( [ eff - ls ] ) to the fundamental " parameters @xmath22 of the quark - lepton lagrangian ( [ eff - q ] ) we use an approximate method based on the standard on - mass - shell matching condition @xcite @xmath57 where @xmath58 are the on mass - shell scalar meson states .
we solve equation ( [ match ] ) using the quark current matrix elements @xmath59 here @xmath60 , @xmath61 and @xmath62 are the scalar - meson four - momentum , mass and decay constant , respectively .
the quark operators in eq .
( [ mat - el2 ] ) are taken at @xmath63 . in the numerical calculations we use the following values of scalar meson masses @xcite : @xmath64 the coupling constants @xmath62 in eqs .
( [ mat - el2 ] ) can be estimated using the linear @xmath65-model in the case of the @xmath50 meson @xcite and by qcd sum rules in the case of the @xmath66 @xcite . in the linear @xmath65-model
one has the following relationship @xcite : @xmath67 where @xmath68 is the number of quark colors . comparing eqs .
( [ mat - el2 ] ) and ( [ mat - el3 ] ) we get @xmath69 the coupling constant @xmath70 of the neutral @xmath51 meson is related to the coupling constant @xmath71 of the charged @xmath72 state due to isospin invariance : @xmath73 with the definition @xmath74 .
the value of @xmath71 was estimated in ref .
@xcite : @xmath75 combining eqs .
( [ apm ] ) and ( [ a0-apm ] ) we have @xmath76 solving eq .
( [ match ] ) with the help of eqs .
( [ mat - el2 ] ) , we obtain the expressions for the coefficients @xmath53 of the lepton - meson lagrangian ( [ eff - ls ] ) in terms of the generic lfv parameters @xmath22 of the initial ( [ eff - q ] ) lepton - quark effective lagrangian : @xmath77 where @xmath78 and @xmath26 . for our analysis
we also need the effective lagrangian describing the interactions of the scalar mesons with nucleons . we take it in the following form : @xmath79 n\ , .
\end{aligned}\ ] ] for the meson - nucleon couplings @xmath80 we adopt the central values @xmath81 used in phenomenological description of nucleon - nucleon interactions and recently also calculated in the chiral unitary approach @xcite . since our analysis does not pretend to high accuracy we do not supply the error bars for the meson - nucleon couplings @xmath80 .
now , having specified the interactions of the scalar mesons @xmath82 with leptons and with nucleons we can derive the scalar meson - exchange contributions to the conversion .
this contribution can be expressed in the form of the nucleon - lepton effective lagrangian ( [ eff - n ] ) which arises in second order in the lagrangian @xmath83 and corresponds to the diagram in fig.1b .
we estimate this contribution only for the coherent @xmath23 conversion process . in this case
we disregard all the derivative terms of nucleon and lepton fields . neglecting the kinetic energy of the final nucleus , the muon binding energy and the electron mass
, the square of the momentum transfer @xmath31 to the nucleus has a constant value @xmath84 . in this approximation
the meson propagators convert to @xmath85-functions leading to effective lepton - nucleon contact type operators . comparing them with the corresponding terms in the lagrangian ( [ eff - n ] ) ,
we obtain the scalar meson exchange contribution to the coupling constants of this lagrangian : @xmath86}^{(3 ) } & = & \beta_{a_0}\eta_{bs}^{(3)}\ , \hspace*{1 cm } \alpha_{bs[mem]}^{(0 ) } = \beta_{f_0}\eta_{bs}^{(0)}\end{aligned}\ ] ] with @xmath25 and the coefficients @xmath87 substituting the values of the meson coupling constants and masses we obtain for these coefficients @xmath88 these values should be considered as rough estimates in view of the uncertainties in the scalar meson masses and couplings .
from the lagrangian ( [ eff - n ] ) , following the standard procedure , one can derive the formula for the branching ratio of the coherent @xmath23 conversion . to
leading order in the nonrelativistic reduction the branching ratio takes the form @xcite @xmath89 where @xmath90 are 3-momentum and energy of the outgoing electron , @xmath91 are the nuclear @xmath23 transition matrix elements and @xmath92 is the total rate of the ordinary muon capture .
the factor @xmath93 takes the form @xmath94 in terms of the parameters of the lepton - nucleon effective lagrangian ( [ eff - n ] ) and the nuclear structure factor @xmath95 that is typically small for the experimentally interesting nuclei . the nuclear matrix elements @xmath96 have been calculated in refs .
@xcite for the nuclear targets @xmath97al , @xmath98ti and @xmath99au .
we show their values in table ii together with the experimental values of the total rates @xmath92 of the ordinary muon capture @xcite and the 3-momentum @xmath100 of the outgoing electron . using the quantities from table
ii we find for the dimensionless scalar lepton - nucleon couplings of the lagrangian ( [ eff - n ] ) the following limits @xmath101^{-2 } , \ \ \ \end{aligned}\ ] ] with @xmath102 . here
the scaling factors @xmath103 depend on the target nucleus @xmath104 used in an experiment setting the upper limit @xmath105 on the branching ratio of conversion : @xmath106 .
the numerical values of these factors for the experiments discussed in the introduction have been calculated with the help of table ii and are given in table i. from the limits in eq .
( [ alpha - lim ] ) one can deduce individual limits on the terms contributing to the coefficients @xmath107 , assuming the absence of significant cancellations ( unnatural fine - tuning ) between the different terms . in this way , using eqs .
( [ alpha ] ) , ( [ alpha - s - ex ] ) , we derive constraints on the 4-fermion quark - lepton lfv couplings of the lagrangian ( [ eff - q ] ) for the two considered mechanisms of hadronizations : the direct nucleon mechanism ( dnm ) and the meson - exchange mechanism ( mem ) .
the corresponding limits are shown in table 3 . following the common practice , we presented these limits in terms of the individual mass scales , @xmath108 , of the scalar quark - lepton contact operators in eq .
( [ eff - q ] ) . in the conventional definition @xcite these scales determine the couplings @xmath109 of the 4-fermion contact terms of the form @xmath110 with a fixed @xmath111 .
historically @xcite this definition of @xmath108 originates from substructure models and corresponds to the compositeness scale in the strong coupling regime @xmath112 .
thus , the individual mass scales @xmath108 , introduced in this way , are related to our notations according to the formula : @xmath113 with @xmath114 and @xmath115 .
let us compare our limits for these mass scales with the corresponding limits existing in the literature .
the limits on @xmath116 can be derived from the experimental bounds on the rates of the @xmath117 , @xmath118 decays @xcite . despite the scalar quark current
does not directly contribute to these processes the limits come from the gauge invariance with respect to the sm group which relates the couplings of the scalar and pseudoscalar lepton - quark contact operators .
typical limits from these processes are of @xmath119 . in the future experiments at the lhc
it is also planned to set limits on the mass scales of various contact quark - lepton interactions from the measurement of drell - yan cross sections in the high dilepton mass region @xcite . in this case typical expected limits for the scales of the lepton flavor diagonal contact terms are @xmath120 .
we are not aware of the corresponding analysis for the limits on the lfv scales , like @xmath116 , expected from the experiments planned at the lhc .
however , one may expect these limits to be significantly stronger ( typically by an order of magnitude ) than the above cited limits for the lepton flavor diagonal mass scales @xmath121 .
this is motivated by the fact that , usually a flavor diagonal process have much more sm background than the lfv processes .
a comparison of the above limits with the limits in table iii , extracted from , shows that the latter are more stringent .
the following comment is in order . as follows from eq .
( [ rme.1 ] ) and table ii , the contribution of the isovector couplings @xmath122 to the conversion rate , eqs .
( [ rme ] ) and ( [ rme.1 ] ) , is suppressed by a small nuclear factor @xmath123 . on the other hand
, the information on the isovector couplings may be important for the phenomenology of the physics beyond the sm allowing one to distinguish the contribution of d and u quarks . in this respect ,
the contribution of the scalar mesons is important , since it is comparable on magnitude with the dnm mechanism . taking this contribution into account we extracted reasonable constraints on mass scales of the isovector quark - lepton contact terms @xmath124 presented in table 3 .
the strange quark contributions are not affected by the meson exchange because we neglected the possible strange quark component of the scalar mesons .
we analyzed the nuclear @xmath0 conversion in the general framework the effective lagrangian approach without referring to any specific realization of the physics beyond the standard model responsible for lepton flavor violation .
the two mechanisms of the hadronization of the underlying effective quark - lepton lfv lagrangian have been studied : the direct nucleon ( dnm ) and the scalar meson - exchange ( mem ) mechanisms .
we showed that the scalar meson - exchange contribution is comparable on magnitude with the dnm mechanism and , therefore , can modify the limits on the lfv lepton - quark couplings derived on the basis of the conventional direct nucleon mechanism .
these results lead us to the conclusion that the meson exchange mechanism may have an appreciable impact on the phenomenology of the lfv physics beyond the standard model and , therefore , should be taken in to account in the analysis of the lfv effects in hadronic and nuclear semileptonic processes . from the experimental upper bounds on the conversion rate we extracted the lower limits on the mass scales of the lfv lepton - quark contact terms involved in this process and showed that they are more stringent than the similar limits existing in the literature .
this work was supported by the fondecyt projects 1030244 , 1030355 , by the dfg under contracts fa67/25 - 3 , grk683 .
this research is also part of the eu integrated infrastructure initiative hadronphysics project under contract number rii3-ct-2004 - 506078 and president grant of russia `` scientific schools '' no.1743.2003 .
99 t. s. kosmas , g. k. leontaris and j. d. vergados , prog .
phys . * 33 * , 397 ( 1994 ) [ arxiv : hep - ph/9312217 ] .
marciano , lepton flavor violation , summary and perspectives , summary talk in the conference on `` new initiatives in lepton flavor violation and neutrino oscillations with very intense muon and neutrino beams '' , honolulu - hawaii , usa , oct . 2 - 6 , 20002 , river edge , nj : world scientific , 2000 ( see also @xmath125 ) ; y. kuno and y. okada , rev .
phys . * 73 * , 151 ( 2001 ) [ arxiv : hep - ph/9909265 ] .
w. honecker _ et al .
_ [ sindrum ii collaboration ] , phys . rev .
lett . * 76 * , 200 ( 1996 ) ; p. wintz , in status of muon to electron conversion at psi , invited talk at @xcite ; a. van der schaaf , private communication .
w. molzon , the improved tests of muon and electron flavor symmetry in muon processes , spring .
* 163 * , 105 ( 2000 ) .
j. sculli , the meco experiment , invited talk at @xcite .
y. kuno , the prism : beam - experiments , invited talk at @xcite .
a. faessler , t. gutsche , s. kovalenko , v. e. lyubovitskij , i. schmidt and f. simkovic , phys .
d * 70 * ( 2004 ) 055008 [ arxiv : hep - ph/0405164 ] .
t. s. kosmas , s. kovalenko and i. schmidt , phys .
b * 511 * , 203 ( 2001 ) ; phys .
b * 519 * , 78 ( 2001 ) .
a. faessler , t. s. kosmas , s. kovalenko and j. d. vergados , nucl . phys .
b * 587 * , 25 ( 2000 ) .
a. faessler , t. gutsche , s. kovalenko , v. e. lyubovitskij , i. schmidt and f. simkovic , phys .
b * 590 * ( 2004 ) 57 [ arxiv : hep - ph/0403033 ] . t. inoue , v. e. lyubovitskij , t. gutsche and a. faessler , phys . rev .
c * 69 * ( 2004 ) 035207 [ arxiv : hep - ph/0311275 ] .
j. gasser and h. leutwyler , phys .
rept . * 87 * , 77 ( 1982 ) .
j. gasser , h. leutwyler and m. e. sainio , phys .
b * 253 * ( 1991 ) 252 , 260 .
j. gasser and m. e. sainio , arxiv : hep - ph/0002283 ; m. e. sainio , pin newslett .
* 16 * , 138 ( 2002 ) [ arxiv : hep - ph/0110413 ] .
a. corsetti and p. nath , phys .
125010 ( 2001 ) [ arxiv : hep - ph/0003186 ] . s. eidelman _ et al . _
[ particle data group collaboration ] , phys .
b * 592 * , 1 ( 2004 ) .
g. colangelo , j. gasser and h. leutwyler , nucl .
b * 603 * ( 2001 ) 125 [ arxiv : hep - ph/0103088 ] ; j. a. oller , e. oset and j. r. pelaez , phys .
* 80 * , 3452 ( 1998 ) [ arxiv : hep - ph/9803242 ] .
e. oset , h. toki , m. mizobe and t. t. takahashi , prog .
phys . *
103 * , 351 ( 2000 ) [ arxiv : nucl - th/0011008 ] .
a. faessler , s. kovalenko , f. simkovic and j. schwieger , phys .
lett . * 78 * , 183 ( 1997 ) .
r. delbourgo and m. d. scadron , mod .
lett . a * 10 * , 251 ( 1995 ) [ arxiv : hep - ph/9910242 ] ; r. delbourgo , m. d. scadron and a. a. rawlinson , mod .
a * 13 * , 1893 ( 1998 ) [ arxiv : hep - ph/9807505 ] .
k. maltman , phys .
b * 462 * , 14 ( 1999 ) [ arxiv : hep - ph/9906267 ] .
t. suzuki , d. f. measday and j. p. roalsvig , phys .
c * 35 * , 2212 ( 1987 ) .
e. eichten , k. d. lane and m. e. peskin , phys .
lett . * 50 * , 811 ( 1983 ) ; v. d. barger , k. cheung , k. hagiwara and d. zeppenfeld , phys .
b * 404 * , 147 ( 1997 ) [ arxiv : hep - ph/9703311 ] ; phys .
d * 57 * , 391 ( 1998 ) [ arxiv : hep - ph/9707412 ] .
j. e. kim , p. ko and d. g. lee , phys .
d * 56 * , 100 ( 1997 ) [ arxiv : hep - ph/9701381 ] .
n. v. krasnikov and v. a. matveev , phys .
usp . * 47 * , 643 ( 2004 ) [ usp .
nauk * 174 * , 697 ( 2004 ) ] [ arxiv : hep - ph/0309200 ] .
a. k. gupta , n. k. mondal and s. raychaudhuri , arxiv : hep - ph/9904234 . | we study the nuclear @xmath0 conversion in the general framework of the effective lagrangian approach without referring to any specific realization of the physics beyond the standard model ( sm ) responsible for lepton flavor violation ( lfv ) .
we analyze the role of scalar meson exchange between the lepton and nucleon currents and show its relevance for the coherent channel of conversion .
we show that this mechanism introduces modifications in the predicted conversion rates in comparison with the conventional direct nucleon mechanism , based on the contact type interactions of the nucleon currents with the lfv leptonic current .
we derive from the experimental data lower limits on the mass scales of the generic lfv lepton - quark contact terms and demonstrate that they are more stringent than the similar limits existing in the literature .
.3 cm _ pacs : _
12.60.-i , 11.30.er , 11.30.fs , 13.10.+q , 23.40.bw _ keywords : _ lepton flavor violation , @xmath1 conversion in nuclei , scalar mesons , hadronization , physics beyond the standard model . |
Subsets and Splits
Data Mining Abstracts
This query searches for and retrieves 2000 documents containing the phrase 'data mining' in the abstract, providing limited but specific document samples.